#opal-general・✨
1 messages · Page 2 of 1
hey all - this channel is for Opal discussions etc, please keep this space on-topic 🙂
🤷
interesting you here?
@bleak anchor I sure could use your help. I'm trying to extract data from a Google Sheet (the asset) and want Opal to make it into a dashboard. Presently, it's creating "mock" data, but I'd like to make it live with real data. Better yet, attach the dashboard directly to the google sheet with a sheets API. Can you direct me to any instructional videos by teachers your @googleLabs team trusts?
Want to transform raw data into an interactive, visually appealing dashboard in Google Sheets? In this step-by-step tutorial, I’ll show you how to build a dynamic dashboard with charts, filters, and interactive elements—no coding required!
Try @Taskade today! https://www.taskade.com/
Practice File: https://docs.google.com/spreadsheets/d/1g...
from my analysis this will solve 80%
You're going to need better data handling. Right now it's basically just reading and spitting it back. You’ll want to structure and clean the data before feeding it in. That way Opal can interpret it properly. You might also need to add a preprocessor node in your flow before the dashboard phase to ensure consistency.
awesome work!
you only have 2 blocks to do all this?
the new tool proves my point
I'll be sure to review the Youtube video, too. This is my first experiment with "vibe coding" so I'm grateful everyone's assistance.
I'm already available
Listened to your audio assist (thank you). I'll I want the dashboard to do is reflect the data for leadership to make easy analysis, so it's ok if visual dash reflects data.
I'll show you two prompts, which were refined on my opal, and were tested on stak
and simply amazing
Is anyone else getting this error in Opac?
"No text generated. Opal has limited quota and it was exceeded. Please try again later."
we might be experiencing some capacity issues
did you know there's a #stax・🥞 channel?
Yes, and because it is in the same refinement flow as the Opal
Yes, I am as well. Was working fine this weekend and earlier this week when I had some free time to build. I haven't made any changes to my flows but was just trying to run it today to show someone what I had built and kept getting that same message.
Hopefully it will be cleared soon
it was exactly the same error as me
several failures
even with the API active
several and several
having an error even in the session
simply move the user
When Lithuania will be in list of early access? @quaint eagle
When Lithuania will be in list of early
📢 We are currently experiencing capacity issues with Gemini 2.5 Pro model. If you see "Quota Exhausted" error, change your steps to use Gemini 2.5 Flash
I use 2.5 Flash in my Gemini app. Speedy and more than capable for most stuff.
I hope Opal gets available in my region , cant wait to try it.
What are some of the innovative ways you all have been using opal so far? I'd like to see some cool examples
Today you will only have this one
Mine is all good now, but thanks for the work around 👍
yo guys
what's the solution for this??
You've used up your video quota.
Are you generating inside the opal?
yes
is there any way to add credits for this?
🌞 all capacity issues should be resolved now
not at the moment, though we're working on making this possible
## Opal Survey: Paid Opportunity
We’d love to get to know more about you and your thoughts on Opal. if you'd like to take our 10-15 min survey and share feedback, you'll receive a $15 gift card from Tremendous (our thank you gift card distributor) for completing this survey. their online catalogs include gift cards and a variety of retailers.
✏️ Take the survey here ➡️ https://forms.gle/tyx97tiFUzY1GLaX8
📣 Click the ✨ emoji below to receive the Opal role, which we'll ping for future updates and Opal surveys.
thanks for continuing to share your experience with Opal here in this channel! 
I think I've found a bug with the Opal output as a "Save as Google Docs" when it is an article and it creates the first Google doc in your Drive, if you then try and run the same (or different) Opal flow again, for some reason it's using the same Google doc as what was created before however ... it doesn't actually update it!
I'm not sure if anyone here is having the same problem or if it has been already reported, but as a work around you have to just export it as plain text
yes, in which case the duplication just doesn't make the change
this will fill the driver quickly if you are not careful
there are 16 agents, multiple working for refinement
I'd rather have an option to save it under a new file, or select which folder the doc is saved into, ideally with the option to select it's filename
No, but in this case, you need to extend it to include that part, you know? It’s not that it already exists it can exist, but you have to build it into the flow.
Do you have the link?
Also the Download File button doesn't work
there is no "download" - it is an IP block for the country.
any work around?
#wait
made a resume feedback generator
pretty good, wish i could output the code and edit it in cursor !
https://opal.withgoogle.com/?flow=drive:/1tuEJl3xKZwpKm0kDOlPZFe0MIkI83e46&shared&mode=app
recommend adding "unemployed. replaced by an Opal." as one of the choices
Not available in our Region
Hey im using opal and hitting quota issues for a workflow but google cloud wont increase my quota. I am an ultra user, can anyone help with this?
whats opal actually about 🤔
-# anyways is there any significant diff between https://github.com/Tencent/Youtu-agent and crewai?
can we have loops
oh wait
LMAO
I mean can we have more efficient loops
can we have json handling too
I take that back
not efficient loops,
we need loops
please
created a few apps with opal today. to confirm, it will always download the output as web file?
Opal doesn't have loops or options. For example, I have a prompt that asks if I want to make an image or a video, but it doesn't work; it does both. The only thing I can do with a little logic is change the input and the behavior of the AI according to that option.
You can have Gemini create a website with your outputs using the output option, and you can also save it to Google Docs.
But I still don't understand something. How Google Flow gives you 8 videos every 2-3 days and Opal gives you 10 per day?
Opal doesn't have loops or options. For
created a few apps with opal today. to
whats opal actually about 🤔
Hey im using opal and hitting quota
Will there be an opportunity to open it to countries outside the United States? For example, Taiwan, thank you
i use windscribe vpn
me too
In the assets, I have a text that says 'Please click on the link https://princepatni.com" after my blog post in order to drive users to see other blogs. It displays the text and also converts this into URL but when I click on it either nothing happens or it redirects to the opal but not to the website
Hello! Is this maintenance right now? I can't access the page, a few seconds ago I could
it says :
It has not been possible to open this app. Please ask the author to check that the app was published successfully and then try again.
It's working again ! 🙂
lele
Hey
what?
opal cant fetch Veo3 right now
This happens when you go to the Opal screen. You know how there are two login buttons? One is in the corner of the introduction video, and the other is at the top. You need to click the one in front, the black button.
If you log in with Google using that other option, it doesn’t save your previous sessions. I lost a lot of projects myself because of this, as they don’t get saved. It works as if it were a temporary account.
So, whenever you log into your account, always check if in the corner of the screen there’s an icon with the initial of your name. When you click it, it should show your email or your name. If it doesn’t, then your project won’t be saved.
I may answer questions in audio format.
For example, when you log into Opal using Google, it doesn’t save that session. You can create flows and projects there, but they won’t be saved. I say this from personal experience two of the projects I had and even shared here were lost because of this. That’s when I realised the issue.
So, every time you log into Opal, to make sure your flow and project are being saved, always check if there’s an icon in the corner of the screen. It should be an icon with the initial of your name or your Gmail. I think it shows the name, actually.
Always pay attention to that detail. Check if the icon is there in the corner, just like in a normal Google account, with your initial. If it isn’t showing, your work won’t be saved.
2.0
A crystal apple levitating above a wooden table, slowly spinning with golden sparkles, then sliced in half mid-air with a glowing knife, ASMR sound of slicing glass.
video request?
🫢👀
I will make a connected flow with two tools
😅
using opal and stax
I thought I would be affected by my current depression situation, but this is stronger
I love doing this
and now literally connecting two tools, for a more than precise refinement
## Opal Survey: Paid Opportunity
We’d love to get to know more about you and your thoughts on Opal. if you'd like to take our 10-15 min survey and share feedback, you'll receive a $15 gift card from Tremendous (our thank you gift card distributor) for completing this survey. their online catalogs include gift cards and a variety of retailers. if you've already filled this out, you don't need to do it again.
✏️ Take the survey here ➡️ https://forms.gle/tyx97tiFUzY1GLaX8
thanks for continuing to share your experience with Opal here in this channel!
<@&1410743137319522395>
成為這個頻道的會員並獲得專屬福利:
https://www.youtube.com/channel/UClPN2rjY4im2LC9vG3Y8vkg/join
推薦共享帳號平台 FlixSeek,除了能使用 Youtube Premium、Disney+ 等影音頻道共外,還有AI服務,例如ChatGPT、Perplexity、Canva也能共享,最近更加入Adobe
影片說明:https://youtu.be/bqRz1e1Ke2I...
thanks for the survey, I enjoyed writing a little about my experience as a newbie who just discovered Opal yesterday! I have been excited to learn more and get better at making mini apps
Heyy I'm from India and it says it's not available for your country, is there any way I can try opal ??
Windscribe vpn
Its in playstore
And in the Chrome extensions (=
Thanks you for all opal (=
Excuse me, is posible to do a question?
Loving Opal, great prototyping tool!
A couple of feature suggestions:
Consider executing parallel branches, well, in parallel. E.g., I tried creating a workflow with 4 deepresearch nodes on different topics, that were subsequently summarized, and even though they were independent, they ran sequentially - and that made whole workflow painfully slow.
Ability to run custom (possibly llm-generated) code in sandboxed environment - similar to ai studio code execution or Gemini canvas
Feedback loops - e.g. creating a critic node feeding back into previous node (until critic finds no possible improvements, with max iterations to prevent halting)
Structured outputs support
Integrations and custom external API support
No doubt team already thought of these and hard at work implementing some, but thought I'd share my view
Hi
I can help you
I dont a staff
In opal you can made ai agents
Whitout code
Its posible do you think is difficilt
First you need to enter in
https://opal.withgoogle.com
And you need to live in usa
Or use a vpn
opal... how would one take a 3d model (step, STL, 3mf etc... ) render it in the best natural position witha background, then tag it with mitiple keywords. package it up with json data and zip file... or am I not gettign the concept of Opal.
how to receive?
there is a spot for you to put your email.
I'll try, thanks
Hello (=
I am doing a hyperrealistic video ai
A its so good
Its posible than you prove it?
When is opal coming to the uk
I forgot to mention this in my survey but it'd be great if we could look back at our edits to our app and rollback to an earlier version in case we keep running into errors. not sure if that's a thing I'm a newbie.
A cinematic one. 
opal is sadly not available in my country 😭
And I present to you my newest refinement flow for ASMR prompts, where it contains an organised structure of content and mental triggers, of mental stimulation, where I combined the three main aspects in this area of digital addiction and intelligent design, bringing together the three best elements into this flow. As a result, it gives me a final refined prompt with precision and a solid structure, where I can control absolutely everything I want.
In total, there are 13 agents. Within the organisation, there are three main ones, which handle the core content, while the rest are divided to carry out the teaching process. At the end, there is a specific agent responsible for restructuring the prompt with all the knowledge from the complete flow. This ensures that all the knowledge generated throughout the flow will be contained in the final prompt.
And what is the purpose of this refinement? I want to study, I want to understand how this idea of digital addiction is introduced, created, and embedded within a video, this version of keeping people hooked on the content. I am using ASMR, which is the most popular among them. So my aim with this refinement is to find or produce something similar to this type of digital addiction. I want to create a version of this digital addiction artificially.
Normally this stimulus, this addiction we could say, is not organic, but it is also not artificial. For example, it is introduced, it is placed into videos manually. There are people who do this, who study design and performance, and they are able to manually add these triggers and elements into the content. What I want is to be able to do this artificially. For example, I want to control how this will appear in the video and whether it will be included or not. So I want to create a way to artificially introduce this into any type of prompt and have full control over how I am introducing it.
And the entire flow runs on the 8-second standard. For example, the whole flow is designed so that every result is meant to last only 8 seconds. This gives me a wider margin. And it becomes even more significant considering that, in research I conducted, the standard attention span of people is 2 to 3 seconds. So out of those 8 seconds, I have 2 to 3 seconds to try and artificially introduce these mental triggers the first 2 to 3 seconds of the 8-second clip.
But I have a small barrier that I am trying to figure out how to deal with For example the standard videos when generated are 8 seconds long In the video generation process when the generator is creating the video the introduction of audio sound effects speech and other audio elements is produced within the first 2 to 4 seconds So in the 8 second generation process those 2 to 4 seconds are when the audio of that video is generated
This means I need to do my work without interfering with that process Because if I disrupt it I will end up with a lot of videos where the audio or sound is out of control or it turns into something that sounds like an off sync dubbing like an crazy
This is a project I’ve been trying to build since 2019. Basically, I spent the entire pandemic studying only this, but I couldn’t manage to do it. Even when creating automation pipelines and everything else, I didn’t have the freedom or flexibility. Nowadays, I basically have that. Everything I couldn’t achieve during the pandemic, I managed to do in just two months, mainly thanks to the new tools being released.
This new tool that has just been launched basically proves everything I’ve been saying the refinement of refinement. So, in addition to having a refined prompt, I can literally take that prompt, compare it with another one, and extract from the two an even better result than either of them on their own. In other words, I can refine, I can improve something without any kind of limit. Because whenever I compare one with the other, the result will always be better than both.
What does this new tool do? This new tool literally manages to extract results from prompts or anything else that has already been refined. So, if I create two refined prompts and the quality and content of the generated video are excellent, I can take those two, compare them in another model, and extract something even better by comparing the two.
This way, I can take aspects from both prompts and generate a new prompt that combines the quality of both into an improved model. So I can extract a better result from the two prompts I already had. Basically, this is unlimited. I can, in essence, improve and refine anything.
@coarse dagger Here is the beginning
I don't want to give spoilers, but you know This bolt.new thing?
⚠️I'm not promoting any tool, I'm just explaining how my flow will do it.
👀
I'm editing the demo video, showing it working completely.😶
⚠️I'm not promoting any tool, I'm just explaining how my flow will do it.
I proudly present my latest project a brand-new refinement flow for creating any type of system or website, powered by Bolt.new. With this tool you can optimize your ideas, generate better prompts, and build smarter solutions faster and more efficiently than ever before.
Just a quick disclaimer⚠️ I am not promoting or advertising any external tools or platforms. This is solely a demonstration of my own workflow and project.
In this request, I asked for a website where I can submit, save, and rate the quality of my own prompts. I wanted the main page to show a dynamic chart comparing the quality of all my saved prompts, and to highlight which one is currently my best. I also wanted the system to let me review, edit, and organize my entire prompt library over time, so I can monitor my progress and keep improving my prompt engineering skills.
@coarse dagger
Just a quick disclaimer⚠️ I am not promoting or advertising any external tools or platforms. This is solely a demonstration of my own workflow and project.
Explaining this flow to you, it contains two agents who are specialists where all their content is based on structures, on commands, and on a specialist in no-code flow building, where you don’t need to know how to program to create something. So, all of their knowledge is based on that. And their knowledge is refined, focused only on this. So, the result it generates in the end is the result that a professional would make from a draft of your idea.
So the flow in its essence tends to understand what you want and build a prompt of what you need. It understands exactly what your request is and what you want to do with the draft of your idea. It’s that, with just a few lines or even with the draft of an idea you have, you can create a system, a complete site. Because it takes that draft and turns it into an idea and in the end delivers you a complete prompt to build that idea of yours from scratch.
Stream spoiler👀
Well, I see that this is no longer my place, and I see that this is not what I wanted, so goodbye.

we're working on it!
Loving Opal, great prototyping tool!
Is it possible to add a cover file for an Opal project? <@&1173646484311384114>
did anyone receive their gift card yet? I just get antsy and worry i typed my email in wrong haha. also thank you to Megan for your new comment in the thread where I was unsure about how to revert to previous versions. that is such a lifesaver!
Then move the image as if it were a movie.
Something I've encountered a couple of times now are long running jobs as a result of the prompting not allowing for a solid halting point. In the future, it could be helpful to have a hard cut off time for a task.
Hey, can somone explain me how to use opal to trigger another opal workflow/app?
I would like to nest some functions. There are some more complex steps/apps that need prepared data, this data gets prepared/editted in a general opal-app. I now want to input the output of the data-app into the complex-app. How do i do that? I didnt found something like a webhook?
Hi
How do signup for Opal? It says 'Opal is not available in your country yet"
I'm in UK
Hey, same problem as above in Germany
Hi all! Opal is US-only for now, we're working to make it available in more countries. Can't give you any specific timelines 😢
Hey, can somone explain me how to use
Something I've encountered a couple of
Is it possible to add a cover file for
It’s not like you can just make your internet act as if you were in the US 👀 right?
I’m trying to test the audio generation model. Input script ✅ -> Gemini Cleans it and improves formatting ✅ -> Generate Audio With ‘AudioLM’ ❌
Error: {"code": "INTERNAL", "message": " An unspecified error occurred while calling the model. Please try again.", "details": ""}
(Not a very helpful error message lol)
Any idea when its gonna roll out to Japan?
Why isn't it available in india?
waiting for it in Canada!
The apps are vanishing in Opal ?
I had created landing page optimizer aaaaand its gone now !
Hi
Primary Action: A powerful, fizzy flow of red and orange foam erupting and cascading down the slopes Foam Details: Vibrant color, with glistening bubbles Lighting: Dramatic, low-key studio lighting that highlights texture and motion Cinematic Technique: Shallow depth of field Technical Specs: 4K resolution Mood/Goal: Epic, awe-inspiring, capturing the raw energy and beautiful chaos of the reaction
Can you tell me until when this survey will expire
I had created landing page optimizer
Hello
Hi
Hi there I hope you're all doing good !
I have used Opal a lot lately.
I optimized my app with many sub steps and feedback loops, I don't know how "heavy" I can grow my flow.
Do you guys have recommandations and best practice guide on Opal ?
Also I have noticed some reccurring errors, especially when it comes to uploading a file.
=> We don't see any upload loading bar which makes it difficult to know if the document was well uploaded
=> Once uploaded I always get an error message saying it couldn't be uploaded. But it worked...
=> Everytime I upload a PDF I'm asked to upload it twice.
=> When using Gemini 2.5 Pro in the workflow the "Generation" output usually includes the reasoning text. So I have to add a sub-step with the Output and then exploit this output in the flow.
I'll try to dig into this channel in order to compile some good practices and answers. However if you guys have advices / answers I'd be happy to have them.
Hi there I hope you're all doing good !
Are there formalities, technicalities or political issues? How can we help?
Hello again. Do we have a subchannel were we share our opals (so lazy people like me can duplicate them)?
hello friends! Question for you: I am using gemini to set up a workflow here and it keeps referring to a text generation vs. code execution that i have to switch in order to create tasks using python. Is it hallucinating? not sure where this setting is. thanks for your help in advnace!
Also any chance there is such a thing as office hours? I'm trying to build an intelligence tool (for the non profit world) and i keep hitting token limitations or other roadblocks and there's op to simplify.
we only have this singular #opal-general・✨ channel! you're welcome to share them here 🙂 and I think some others have shared above too
this is a sweet idea 🙂 will chat with Opal crew to see how we can help!
it's work in progress! primarily process related but also resourcing to make sure we can support more users in other geos. To start, we'd also only support english since there's more work we need to do to localize the prodcut and our natural language prompt generation/editing flow.
It also helps to know which countries are most in demand! So please feel free to comment or leave emojis with your country flag if you're waiting for Opal to come to your country 🙂
I haven’t received the payment for the survey yet. Can you tell me when I will get it?
Anyone have any guesses as to why I can't ever get to the Output phase? I'm trying to build an intelligence report (for the social impact community) and the info is actually brilliant. It just happens to fail at the HTML output stage. Feedback: I wish there was a feature to retrigger a single step (like N8N for eg.) after failure as it makes testing easier! I have to wait like 15-20mins every time to restart and test the flow.
Anyone have any guesses as to why I can'
When will Opal launch in India?
Same question from Taiwan
Hey everyone! I'm guessing I'm not the first to ask the question, so apologies for any repeat. I'm just starting with Opal...currently, the trigger for action appears to only be user input/text, right? Curiuos to see if API/Webhooks are in the works (or if I'm overlooking it)
I saw an n8n Post about automating image generation from multiple sources.
I developed this as a parallel example in Opal.
Ahem... It took me only 5-MINUTES to develop in Opal... with ZERO-SHOT perfect (IMHO) results!
Kudos to @quaint eagle and the reset of the team! Very cool!
I think that is all there is now. I'm wondering if there's a feature voting list somewhere, or even the ability to read the source, or open source it like the gemini-cli.
I would love to add an MCP block and/or A2A block if there was a way to help out at that level.
Triggers are on the roadmap!
Opal is built on top of the Breadboard AI project. You can follow the development here: https://github.com/breadboard-ai/breadboard
A library for prototyping generative AI applications. - breadboard-ai/breadboard
where is the roadmap? 🤔
I would be OK with a "trusted" view/contributor.... afterall the world is full of snakes trying to out run each other on roadmaps.
You guys have done great with it so far! I had a set of gems I used that I would copy and paste between and I had a workflow document outlining them all. I just copied and pasted the entire 4 page dovcument into the prompt and voila! Well done!
@fading pulsar you might consider a search node to generate some ideas for relevant topics on a story idea to pick from before jumping into executions. (just a thought - could not tell if you had that in your workflow or if you could even do that as a step)
Great idea! I have a long list of ideas in a separate doc that I've built over time, though that would be a great addition.
yea - I am finding a little market research with AI is not a bad thing to get some other things moving on the idea board.
OMG that is great - is that available to remix?
Not available in Pakistan. Is there any way I can get an invite to test this?
I am getting an error when i try to use AudioLM Model in a node to generate speech from text. opal workflow comes till the speech to text node and then i get an error saying - unable to fetch the output. i also see this warning-Audio generation has limited free quota. i have subscribed to AI PRO plan. not sure how to fix this issue. pls help
Waiting for "Looper" 🙂
Opal does not support - Create an app to convert uploaded PDF or Doc/Docx file to text. Is there a way to take document from Google Drive to process?
Cool
Hii
how many time to repair in the free trial
are their instructions on how to connect Opal apps to my google drive/email accounts for workflow automations?
Hi how can we use a paid api key within opal? I'm trying to build apps to generate videos in sequence but i keep hitting rate limits -- im willing to pay to get rid of them i juts dont see an easy way to do this
When will Opal be accessible in Pakistan?
Building a quick meeting note taker from video and getting "Internal Server Error" due to the request is too large. Any way to 1) increase the timeout of the server to prevent timeout 2) Increase the token size of the LLM to allow it to generate larger response
Hi how can we use a paid api key within
Building a quick meeting note taker from
I am getting an error when i try to use
we don't have timelines for specific countries, but we promise we're working hard on it! please continue to vote to help us prioritize the countries you're in 🙂
Google drive is native already, so you can upload or save to drive! Juse use the @ symbol when you type in your step to see your options. We're still working on gmail integration, but eventually that should be just as easy!
such a great use case! sounds like you have some followers if you decide to share the Opal 😄

whens the global release i dont wanna use make.com anymore
oh lol i just saw the rest. fr tho i hope we can get it before christmas 
Could you please provide more details on the steps that led to this error, as we wish to investigate it further?
:JoePain:
When will Opal be accessible in Pakistan
how many time to repair in the free
Something went wrong. Please try again or send feedback.
Something went wrong. Please try again or send feedback.#opal
Is anyone availabel here who can provide me understanding to know "how I can use Opal in Project management domain"?
Something went wrong. Please try again or send feedback. #opal-general・✨
Let us use #opal-general・✨ from LATAM and we can help we so much feedback 🙏 to enhancing the tool.
Hi
I don't know how to tell you the details.
What is the daily limit for the number of videos that can be generated by a Veo3 node in Opal?
Google afternoon, please can we générate an one minute vidéo with opal?
Imagine that! After I saw your Flow
No, because there is no node for video editing and splicing
I'm running into a similar issue. I have AI Pro Plan and I haven't generated a single video today and this was the first Opal I ran today as well. Yesterday, I ran into the same issue as well but, oddly enough, I was able to generate a video clip on the second attempt with my Opal.
How do i get access to opal from Ghana
Man, if they allowed that I believe we'd have some solid short films and some YouTube kids bait from hell asap on Googles dime
With sonic and knuckles of course
Opal is currently available in the U.S. and is expected to expand to additional regions. We will keep you updated once it is available
Are there any opal docs that explain the different asset types? For example how to connect different assets into the flow? For example -- I have a finely tuned deep research prompt I want opal to execute, I don't need opal to generate a new prompt, just insert the user inputs. I thought the text asset could be useful, but not sure.
Hi
Can 10 Veo3 models be run concurrently on the same Opal node?
Hello
I'm running into a similar issue. I have
Also curious about the code execution element and how to use it.
we're working on documentation now, so I'll flag this request! To answer your question though, you just want to offer a user input that feeds into a deep research prompt, right? In that case, you can just add a "user input" step to the graph and connect it to the deep research prompt. Make sure in your deep research prompt, you tell the model when to use "@user input", you can use the @ sign to pull up a menu and select your user input. If you wanted to provide a document or image that doesn't change (IE a user isn't providing those) for example, then you would add those as "assets" instead of a user input. Does that distinction make sense? Lmk if you have more q's!
Also curious about the code execution
Thank you @elfin tulip -- i actually got it to work today. yesterday when I posted this, I could not get a generate step to reference the text asset. but today it worked. maybe a temporary bug.
New question -- does opal support generating multiple separate output files? I tried the following but it resulted in a single output file. I made a flow to parse a deep reseach document through four different lenses, each "lens" requires a distinct generate step to output the correct output. In my case, I am looking for content in the deep research doc and parsing it into four different structured json files. I suppose a workaround is to have one step create a massive comprehensive json structure, but then I'd need to break it up again.
Yo, anyone know if you can make Opal do an outbound post request to a server?
I have it formatting some JSON, and I have a server set up which can accept JSON,
is it possible to do a POST request from Opal?
I've got a few other ideas of how I can achieve similar functionality, but this functionality would by far be the easiest
I can't imagine Opal not having MCP support at some point. @worldly forge ...
yup, actively working on MCP support
@quaint eagle Will Opal projects be abble to be portable to run-inside a private environment? (MCP seems to be an enabler for that)
@Dimitri Glazkov (Labs) Will Opal
Yo, anyone know if you can make Opal do
New question -- does opal support
no, veo capacity is limited so you would likely get rate limited
Let me look at the error message tho, it looks ugly
I noticed the other labs projects have additional channels for sharing work, but I didn't see that for Opal. Don't want to complicate things, but I would surely love to see what some folks are building, particularly around how they're using the code execution piece.
Potential bug: I set up a simple flow with 4 user inputs that are inserted into a deep research step (TY @elfin tulip ) and output to google doc.
I ran it once with inputs for "regie.ai". Then a second time with inputs for "observe.ai", however the outputted google doc did not change. see screen shots.
I suggest people to take a trip to the USA quickly , once in USA you could enjoy Opal or something else who know !
think that a lab purely experimental non dangerous and often very interesting may justify the trip 🤫
I'm all ears
Will it support 9:16?
Ok.thank you
quickest and simplest to travel for free and safely to the USA would be to not forget to use Urban VPN or Windscribe in airport better to use a solution like this in browser !
Anyone knows why veo 3 doesnt work 9:16 and img 2 videos are always slow motion
I used a workflow to gather intel from products using my image then to make a profesional production prompt for veo 3 but is alwasy slow motion
Is it just me, or did all your View-Only app shares suddenly start requiring Google sign in and all the phishy sounding permissions for Opal ("See and download all your Google Drive files. Learn more"). All the people I have shared with who were fine using the app up to yesterday are now flooding me with, "What gives? You want to read my entire Google Drive?"
Animation & Camera:
Bedroom interior, morning light through window. Camera slow zoom on boy (10 years old, Arab, white shirt, navy pants folded on chair, still sleeping). Mother enters, animated hand gesture to wake him up. Slight handheld camera shake.
Sound / Music:
Dramatic slow music, alarm clock faint in background, mother’s footsteps.
Voiceover (Arabic):
"اصحى يا حبيبي… المدرسة بدأت!"
Need some feedback with a mini app, i made gemini gerate a ui for uploading images to crop them with 16:9 ar but i cant make that image go as output and input for veo 3, its always the text html
My feedback would be to try to use something that doesn't require your end-users to give their entire Google Drive away to use your app. 😉
you can just use a dummy account with nothing in it
Yeah, but I want real people to be able to use this without the hoops to jump through.
I know what you mean sir
Giving full access to drive its like giving everything for them to scan
Trying n8n now. Was using Replit, but ran out of free access before I got a mvp done... Replit was looking good, though... may return to it and actually pay. 🙀
Hi guys , I want to ask if located in Uae and when I am trying to sign in at Google Opal it's writing that's not avaliable in your country right now, so if some body know how to fix it pls help me with it , THKY
NordVPN?
Went with Replit afterall. App is done and doesn't require end-users to grant all access to their Google Drive. 🙂
We've always required sign-in for running an app, but we did just allow users to view published apps first (only prompted to sign-in when they click run). We also do require Drive permissions to save things to your drive (if you select output as save to doc/sheets for example) or to allow you to upload files from Drive! Can you send a screenshot if anything looks off though? Sorry it looks spammy, definitely a legitimate use case, we promise 🤣
great idea! We'll work on getting one set-up
great idea! We'll work on getting one
I have a question. is there a way to get the opal apps to store memory? I am trying to make one that analysis certain statistical probabilities but it needs to learn from the past data it generated, and improve the accuracy of the reports over time. I am assuming this is not possible yet with opal apps?
analyzes*
Need some feedback with a mini app, i
When is Opal coming to EU-plebs? 😢
not yet possible, but on our roadmap! very cool use case 🙂
Hi guys , I want to ask if located in
Anyone knows why veo 3 doesnt work 9:16
Will it support 9:16?
Getting access blocked on Opal for Advanced Protection.
Yeah... a view-only would be great. Even my girlfriend of 20 years who I asked to help test said something to the effect of, "I am not giving your app these permissions to my drive! ...and, no one else will." Hmm... maybe I have a bigger problem... 😉
Hey everyone! So im kinda new here and i heard a little bit of Opal but i dont really know what is the use cases of this tool. Is it a sort of a no code automation builder?
yeah
Is there a way to make a loop? Or is that not possible yet?
Hey Braaazill, wheeen? 😄
soon, very soon, maybe sooner than you think. 👍
Will there be an option to publish/share as a website so that people wont have to sign into google to try out what we build? Eg: https://opal.withgoogle.com/?flow=drive:/131XojJ3IMbp00E_191jCnM5awt_vzfBP&shared&mode=app
How can I get Opal to output a consistent HTML designed document? I get a different design every time I run the application.
I mean it defeats the purpose because that is what allows everything to work so seamless. Kind of like a whole eco-system that is all automated nice. imagine the possibilities. Plus it's 2025 time to get a FREE google account. But in fairness I'm sure that will be a "premium feature" or "developer mode" something of the sorts. where this would be our back end for the projects and the link would just provide a nice clean "AI" chat interface. maybe use as "guest mode". could be implemented on some type of timer for people to test out the apps. Not sure...
Please consider Peru, in the next release. I would like to test Opal !
Agree with a lot of this! It's hard for us to allow signed-out engagement since we host the published opals and need to enesure quality and prevent abuse . We are working on making the sign-in process more seamless so it's fewer permission clicks for your end users!
not possible yet!
very soon! subscribe to our Opal channel alerts, we'll message when our next wave of countries is out. Not saying anything officially, but pay attention in the next few weeks 😉
We're still working getting a sub-channel to share your work, but in the meantime.... I made an opal to help smaller brands make "exploding box" type video ads. Recorded the video demo and will also share the Opal! Will upload video in a bit, Opal is below
for newcomers like @soft sky , you can check out this example video for more context on things you can do with Opal!
sorry for delay, ironically being blocked from uploading a video to discord right now (thx google security 😆 ). Posted on Linkedin though, so you can view the video there: https://www.linkedin.com/posts/megli103_do-you-remember-the-ikea-exploding-box-ad-activity-7377009329067139073-PJML?utm_source=share&utm_medium=member_desktop&rcm=ACoAAA3fHeIBbPmZLe5JnaUFPgZH9dBQ8HBmnG8
Do you remember the Ikea exploding box ad that went viral? I wanted to see how hard it was to recreate this ad format for other brands, and I was blown away with how easy it was to replicate with Veo3. It was pretty simple to get a high quality video with one prompt when I wanted to make a video for a national brand like Home Depot. When I tried...
Hi all, I'm building rather large apps on Opal (loving it so far), but I started running into two issues:
- The elements of the canvas start flickering and disappear as well as the text in the side panel, when I zoom out to see the whole app.
- The apps stops at some point and throw either of the following errors:
a. "Request had invalid authentication credentials. Expected OAuth 2 access token, login cookie or other valid authentication credential."
b. "HTTP error: 504, details: upstream request timeout"
Any idea what might be happening there? Maybe a limit of how many calls I can make within a given app?
when I try to use veo 3 makes the same thing for me.
Interesting. So I switched everything from Gemini 2.5 Pro to Flash and the app ran. So looks like I hit a usage limit for 2.5 Pro...
It would be great if I could simply pay for usage, rather than having the app stop running at some random moment. That puts me into a complicated situation with having to demo a prototype tomorrow.
Now it seems I hit the Flash limit as well. Getting an error: Failed to fetch because one step doesn't run anymore. 🤷♂️
Interesting. So I switched everything
Hi all, I'm building rather large apps
hmm... I wonder if Pro is having availability issues. They had a pretty big launch today
I change everything and always this message "Something went wrong. Please try again or send feedback.
" why ?
One more thing, the "send feedback" button is loading forever!
I am from EU if I VPN to US - it wil work?
yeah. you will have a lot of errors.
Guys I cant enter to opal, I try many times but there's nothing
/image-to-veo
How I enter opal from discord guys need help!
I cant enter to the site
I try many times but nothing happened
what message are you getting when you try? Could you describe the steps you're taking?
I put the email & mdp and then I get back to the page
Can I make vedeo veo3 from here I really need help
When the opal will be accesable in Poland
When can other areas use? Would be great if it is soon. Im so curious ^^
just use VPN
Anyway to disable remixing?
I have a Google AI Ultra subscription, and I want to generate unlimited VEO 3 fast videos in Opal. What can I do?
How to use opal in Bangladesh Please reply me.😩
use a vpn my friend. but the tool still very basic and google products doesn't work because it's free.
Tnx For Your Advice🙂
the flickering and disappearing UI when zooming out is likely a rendering performance limitation in Opal's canvas handling for very large apps
and the authentication error suggests your session or token is expiring, so refreshing or re-authenticating should resolve it
the 504 upstream timeout usually indicates your app is making too many or too heavy requests at once, hitting server-side or execution limits
how to enable Opal in Google Workspace?
Hi, do you plan adding api calling features or else? Is there a roadmap or something?
you need your IT admin to make sure "experimental/early apps" is enabled! https://workspaceupdates.googleblog.com/2022/07/configure-access-to-experimental-google-workspace-apps.html#:~:text=This feature will be ON,using your Google Workspace account.
we're working on this integration so that we can detect automatically your subscription status. Not possible today, but coming soon!!
not today, but that's an update coming soon 🙂 you can also hide your editor view once we update the publishing flow
When will Opal be available in Romania?
thank you Megan ❤️
Hello, we wanted to remind you of our server guidelines, stated in #start-here・📍 : “Currently, our team can support engagement in English; the discussion in this server will largely remain in English for now.” Please keep in mind that non-English discussions and content may be removed
Hello, i want to connect opal with my business whats app. Can we do that?
not at this moment! out of curiousity, are you wanting to have an opal send a message on whatsapp for you?
I haven't been able access Opal for a while. I am in the US and every time I log in, it goes right back to the start page and I log in endlessly but can no longer access the actual program
I've been having the same issue for weeks
Have you tried another browser? I havent had any issues…. Maybe, hopefully just a cache thing
Hi @elfin tulip any way for users from India to try Opal? If there's any beta program or something else which enables. TIA
Yes! Like an ai agent
Do you know whats the limit for generating videos? I generated 10 of them until i got this error. I thought it was 10 per day but after waiting 24 hours i still got the error.
too much load.
I mean like, is it 10 per week, 10 per month?
10 per lifetime. in your lifetime you only can see 10 videos generations. more than those, it's not good for your health man.
no beta, but we may have an update for you in the next few weeks.... can't share officially but it's coming very soon 
oof, that's a bad experience!! Sorry you can't seem to sign in. Can you share what browser and device you're using? If you're on mobile, I've heard that bug before but hadn't heard anything like this for desktop
quota should reset daily!
@elfin tulip I use chrome on my computer (HP) and phone (Galaxy). It happens on each device.
@elfin tulip I tried from my personal account (Ultra) and from my business account. No luck either way.
Hmm I can check our rate limit set-up but it may be every 24hr instead of every calendar day
🪲 🔨 Hi Opalites, we need your help testing! Next time you're making a new Opal, can you use opal.google URL? You'll need to log out and log back in to see your old opals. Otherwise, everything should work exactly the same is it does in opal.withgoogle.com. If you hit any issues, please tag me. Thanks!
the problem @elfin tulip is that it's not available in all countries except US!
@elfin tulip please hurry and connect my ultra account to opal because I see all these n8n automation videos, and wanted to do it on all google products, tried opal and it wasn't working well so I've subscribed to ultra to use the paid full version. and my subscription will end soon and I will not be able to do long form Veo 3 content automation by opal. thank you.
This worked for me. Never seen a .google domain before! https://opal.google
I am trying to use opal to generate 9:16 videos. Any ideas how to do it? I have tried without success so far. Thanks.
easy
Keeps bringing up how much access Opal already has to Drive and other services every time I log in.
I can't see my old Opal apps even after logging out and back in.
You cant use 9:16 yet its default 16:9 also if you disable prompt expansion it will throw error
I hope they fix all of these problems, I think this is a great tool.
@timber shuttle 👍
image generation with nano banana seems to be way off today. Could there be some sort of regression?
is there a way to use pr export Opal workflows within webapps?
it's block by google
Oh, that would be lovely!
thanks for testing!!
can you try again? We just optimized image gen in Opal over the weekend, let us know if you're still seeing quality issues!
yeah we're working on making sign-in process better as well...I know it sounds sketchy every time they notify you with a "security alert" on sign-in :🫥 I promise Opal is secure!! 😆
oh and we should now be able to see all your old opals (lmk if it's still not showing up for you!)
honestly pretty high, don't expect most users to hit it (can't recall exact # off top of my head)
Thanks!
Hi @elfin tulip, can I DM you about the hybrid community pilot you posted about?
yes!
Done! Feel free to check at your convenience.
Thank you!
image generation with nano banana seems
is there a way to use pr export Opal
What is the cap on image generation with
I am trying to use opal to generate 9:16
@elfin tulip when can I use my ultra account in opal?
hi chat
i just got a lot of problems while trying to use opal
any one could help pleas?
OpenAI has already released Agent Builder; isn’t Opal open to other regions yet?
Good morning
Opal team. With the Agent Builder worldwide launch by your nemesis, what's your next product move?
Great time to be a Product Manager. Both teams keep me on edge!
https://youtu.be/44eFf-tRiSg?si=hiwKtz3nJCHFKOWy
Christina Huang from OpenAI guides you through Agent Builder—a new visual tool to create agentic workflows.
Drag and drop nodes, connect tools, and publish your workflows with ChatKit and the Agents SDK.
Is that's why it won't let me use it?
And is get open by the time?
i just got a lot of problems while
Yes, we were watching that launch! Agent Builder feels a bit more focused on traditional dev-focused building flows. We're hoping to make building a bit simpler and easier to access without as much of a learning curve, so it's easy to iterate and debug without needing to code. Exciting to see this space heat up though!
Stay tuned in the next 2 hours 🙂
Great! Because that's the moat I see in you 💖Opal team.
My Prod×Mng money is on you! 🏇
B•U•T, pay attention to that widget feature they have, 🆗?
@Megan Li (Labs) when can I use my ultra
Major Opal Updates 🚀
Thanks to all your feedback, we’ve just pushed a major update with some of your most requested features. Plus, we’re officially welcoming builders from 15 new countries to our community!
Here’s what’s new:
- Interactive, No-Code Debugging is HERE! 🐛 You can now run your Opal workflow step-by-step and iterate on a single part of your app without running the whole thing. Errors are now highlighted on the exact step where they happen, so you can stop guessing and spend more time building.
- Blazing-Fast Performance ⚡️ We've made massive under-the-hood improvements. In some cases, creating a new Opal can be up to 50x faster (under 100ms!). We also enabled parallel runs, so complex apps with lots of steps will now finish in a fraction of the time.
- Geo Expansion is Live 🌎 Opal is now available in 15 new countries, including India, Japan, Brazil, Canada, and more! We’ve even added localized galleries with Opals relevant to new regions (like a Diwali-themed Opal for our builders in India). Please give a warm welcome to all the new members who will be joining us!
- Fresh Look & New Home ✨ You'll notice a new landing page when you go to the homepage while you're logged out (you can also reach us at opal.google now). You can now upload a custom image for your Opal app in the theme section so you have more control over how your app shows up.
- New #opal-show-n-tell・🖼️ Discord channel 💬 so we can share more of what we're creating with each other! Join this channel to get inspo and share your ideas!
We are so grateful for this community. Your feedback has been critical in shaping these updates and the helping us decide what's next. Check out the full story on our new blog post: Keyword Blog Post
As always, our team hangs out in this channel so don't be shy and say hi! 👋 With gratitude, Dimitri obo the Opal team
PS. we know that's not all the countries you're waiting on! We're working on unlock even more geos in our next wave. Feel free to add your country flag if you're still waiting on Opal!
Also feel free to send us feature requests 🙂 Thanks Opalites! We appreciate you 
Copy/Paste it seems for OpenAI on a bit of that. #xerox #memories
Please!! consider Peru on your next wave, we cannot wait to try this app!!
Congratulations on the expansion! 🙌
to be fair, we launched Opal first 😉
Technically "not launched" - in "labs" -- so there is that... Not to mention they were probably looking closely at the open source predecessor.
I've tried to behave myself 😂 "Inspired by", but it looked very similar. Maybe different scope? Advancement all the same. I'm just happy people are finally excited about something besides model improvement.
@elfin tulip don't get me wrong - I am a huge fan - but we kind of need more 🐄 + 🔔 on releases of cool stuff that others are copying.
Totally agree actually, something we need to do more of! Flagging this with teams internally, thanks for the feedback!!
"Xerox"
Oh, you complete me!
PS. we know that's not all the countries
When will Opal be available in Germany?
There is still no Taiwan.😂
I'm going to go use the Agent Builder.
Hi everyone! How do I download images generated by the Opal graph I made? I don't see the images on the canvas and if I look at them in the Console tab, I can't right click to download. I then tried saving them to a Google doc, but that isn't working. I get "Must specify at least one request." No other error msgs are shown.
Hi, I have a feature request. Please add gemini 2.5 Computer use to opal. Should be so cool if opal can make apps like Project Mariner
oooh yes, we've talked about partnering with them on some kind of integration!
also realized we haven't talked about this,but we finally have some docs to get folks started 😂 https://developers.google.com/opal/
Oh I'm more than excited! Can't wait to try it out
no promises! Still discussing what might make sense
Sure! I'll love whatever progress the product gets
@thorny grotto I had the same thought about computer use + Opal. This would be a major differentiator from n8n and maybe OpenAI's Agent Builder (haven't tried it yet).
I've been playing around with Opal today and I looked at the Console logs. I'm noticing something interesting. It's doing some prompt injection to slightly modify the user input prompt.
Then when it uses the Generate node with Imagen4 it uses 2.5 Flash (which I didn't ask for) and it generates a very long verbose prompt for an image gen. I didn't like the image it output.
I took the prompt my 2.5 Pro node made from the initial user input and used that prompt directly in AI Studio to make an Imagen4 image. It looked much better.
Is there a reason why the nodes are set up like that?
I've been playing around with Opal today
Hey! Is there any chance Opal will be expanded to Google AI Ultra subscribers not in the current Beta regions?
Not likley, but we're trying to expedite Opal access for the rest of the world 🤞 Thanks for the patience!
Would love it. And the fact that opal is free makes it accessible by more people
sup
Hi everyone.. I was hoping to use Opal from Nigeria as a beta tester..
I have experience building products with AI and would really like to test and share my opinions on all things design-related when it comes to opal...
I have some ideas I would like to test out..
Here's to know how do I get a early release preview of Opal.
It currently says
[We are sorry, but you do not have access to Early Access Apps. Please log in to your Admin Console to enable it. Go here to learn more.]
I have a business starter account for Google.
What is this opal i cannot understand
Sooo... How about that? A Google Product is not allowed by Google's own security policies. 🤦🏻♂️
Sooo... How about that? A Google Product
Does anyone have a useful way to obtain an access token via Google API within Opal so I can make API requests? It is able to tap into a specific URL with a POST and the provided code, but I'm stuck at this step.
Does anyone have a useful way to obtain
you guys had Opal first — but of course you kept it only for the U.S., because why should the rest of the world enjoy it, right? 😂 Then suddenly, the moment Google saw OpenAI drop its new killer update, boom — Opal magically rolls out across Europe. Gotta reel in a few more members, huh? Now you’re holding that ace up your sleeve with Veo 3.1… or maybe it’ll flop just like Veo 3 did. Seriously, just stay in the U.S. if you’re not planning to launch anything properly in Europe.
We are planning to launch to Europe! Just have a few more regulatory barriers to clear before we can do that. OAI's agent builder launch is definitely playing in a similar space, but it's positioned more towards a dev audience so it's not quite the same from where we sit 🙂
What is this opal i cannot understand
Just wanted to say. I don't get it. I would get it if it saved some kind of code when you publish a project. But AI has temperature. It never does the same thing twice. And even if it did you would then need to keep exactly the same model running so the app looks the same in 20 years time. It fails two key fundemntal principles in engineering: consistency and reliability. Not only that but it seems to push towards a future where everything on computers is interpreted, not compiled. C++ has proved this idea wrong. No matter how powerful machines get, we always want some things running at maximum speed. Happy to discuss. And of course I could be wrong. But well done for releasing this controversial project.
Just wanted to say. I don't get it. I
Give us the MCP connector 🫣
👏🏾👏🏾👏🏾👏🏾
Hi ! Can I make only by promoting a Chrome browser add-on for accessing some data in a website ?
Thanks !
Hi everyone. Google Opal is available in Brazil!!!
Give us the MCP connector 🫣
hey all I am building an app and I wondered is there some way to keep the UI exactly the same between updates? Each time I change an asset or generative block the app refreshes and makes anywhere from small to large changes to the UI.
sure is 🙂 Opal expanded to [15 new countries](#opal-announcements・📣 message)
🙂
Not right now 🙁 You can save a particular output if you want to keep that version, but I hear you on saving a consistent app output. We're working on a way to maintain consistency and give you more control over changes to output to address exactly this
cool thanks
How long are shared projects/output valid, will they expire eventually?
Can you add a feature so an output can go into something else please? If it's a webpage then it can pass in the html file. Much needed. Then this would be a killer system.
Where can put me list and can we sign up for nda or something try it out in Europe
How long are shared projects/output
Africans watching opal go across an entire ocean to Europe
to be fair, we launched Opal first 😉
can my Opal projects read AND write to sheets in my drive? Can I store user data for example to track wins and losses in my game?
sounds like a no lol I asked it to make an app that stores usernames and passwords and it said that app requires tools that are not currently available
For those of you trying to make apps with Opal, if you go to Chat GPT and ask GPT to make you a script for app(whatever app your making) and whatever you want your app to do. It will make you a script for use with Opal, all you have to do is copy and paste and your app will be made completely without anymore input from you.
when will opal be released in France?
<@&1366831844502274180> When you share your app, does it allow them to change anything on app. With the remix button it seem like they are able to duplicate your work. I think it would cause a lot of issues with the creators of App?
Hi, us mods can only support with issues regarding Discord. Hopefully a Google staff member will see your question and be able to assist
They are often quite active in here so should see your question at some point
these apps are really just experiments/protoyping for mocking up an app and playing around, I think the idea is almost that these are open source projects, if you are making something proprietary, then only share the link with your team.
I'm so close to my app being perfect but the generative part of it makes it nearly impossible to specify exactly what I want lol so I am jsut "rerolling" my app until its how I want or good enough haha
Opal is currently available in 15 countries! As for France—it’s coming! 💃🏾 @elfin tulip said “We are planning to launch to Europe! Just have a few more regulatory barriers to clear before we can do that. 🙂”
<@&1366831844502274180> My share links stopped working, and it says unable to share board. When I try to get a share link all it does is spin and shows nothing?
unfortunately we can only assist in moderation issues. please ping the labs team or open a thread in #1238621511401734254 for more help :)
@quaint eagle I posted a screenshot of each error for you.
Hi there, I posted some Opal content on #todays-special・🥣 yesterday and encountered an error, so I’d like to report it.
It would be cool to turn our opal workflow's into callable API's or something that would be insane
maybe overpowered
I 2nd this request after working on a super simple chat bot loop. An "output-to existing file" node would satisfy this need
Here try this link, its working now. https://drive.google.com/file/d/1u46nH8YdUFwL4_4rWJik2_bIi23e-rwY/view?usp=sharing
👋 Hey guys,
I just built a sample app that analyses your resume against a given job description and provides a detailed report on your chances of getting selected.
I’m sharing it with my community, and it’s getting quite a bit of usage. However, I’m currently unable to track the usage and generated data for research and analysis. Is there a way to track this effectively?
https://opal.withgoogle.com/?flow=drive:/1vK8K_YVu-ciwHsSt5I-cKWGISdMCwIJ_&shared&mode=app
Hello all, why isn't Opal available yet in the United Arab Emirates? Is there any plan to expand the regions soon?
I'm sure they're going to be in every country soon it's just in experimental stage they just mentioned availability in 15 more or 15 total countries my memory fails me but I'm sure that's on the roadmap unless there is some sort of privacy laws I'm just speculating 🤔
isnt this same product as gemini? i sometimes feel like google is behind with giving everyone at once access to test stuff
im not sure what all the stipulations are. Yes it is ran on gemini but there might be some data laws that conflict with your/their country
If you get a solution running, I’d love to hear how you engineered it!
I think we may need to come up with other ways of saving results. Recently, I’ve been playing with chrome extensions. I think the beauty of it is that you can use it without even noticing.
I’d suggest you could engineer a chrome extension to utilize your app and save the info to the db of your choice. Although this isn’t the ideal method (forces users to go on their computer, breaks privacy, etc) it’s a method that could work sharing with close friends until more integrations happen.
Since you mentioned community, another suggestion would be to share it with an incentive. “Try my app, comment a screenshot of the results, and enter to win a dinner with me….on me”
Hope this helps!! 😊
Maybe they doing trabslations😀
Yea that's the beauty of speculation it could literally be anything 🤣 they could have it all ready and be just sticking to a roadmap who knows maybe those other countries have beta testers already 👀 gets the gears going doesn't it?
Just to add to the convo 😄
I think it’s smart to rollout after careful testing. As you may know already, if you prompt with a different language you could get a different output or break limitations that was previously guarded… and that’s just one example of how you could break it haha. With Google being a big company, I’m sure it’s better to be second and watch your competition rather than jump the gun and burn in flames 😅😅.
we use the Gemini models to power Opal, but it's a different standalone product. That's why we still have to evaluate country by country requirements to make sure we're abiding by various regional policy/legal requirements 🙂 But agree that we are hoping to roll out to the rest of the world asap! It's top of our piriorities right now!
+1 to localization testing, that's why we're starting english first and doing a more gradual rollout. Once we have more language evals in, we can roll out a truly localized product per country. In the meantime, we have no language filters, so you're welcome to try natural language prompts in your local language 😉
👋 Hey guys,
thanks a lot! iam now in waiti mode 🙂
You inspired me to use Opal to make a landing page. Have you tried that yet? Might be more interactive! I’ll report back on my efforts.
I am from india opal is not available in india right now
I want to make a marketing agent workflow any best automation website recommendations?
Thank you!
I am from india opal is not available in
Somebody knows how to restore Opal Apps in the folder but not reflected in the Opal Webpage?
Does Opal uses veo 3 quality or 3 fast
You may need to create a duplicate for now… that’s how I got it to work
Eyy
@frosty thicket can some one join me and help me download please
@Al Marks (Google) can some one join me
FYI - we're starting to redirect visits to opal.withgoogle.com to our new URL Opal.google today! This should happen silently with no disruption on your end, but please let us know if you hit any issues 🙏
ok
no problem
Has something changed? I tried to acces it today and it sais it is not available to Brazil?
Has something changed? I tried to acces
Hello Im korean user so i can't text to English well
When I use the "Save to Google Sheets" I got this Error "Unable to parse range: Sheet1"
I think the when we make a new sheet in Korea. First sheet name is "시트1"
so that's not a match "Sheet1" & "시트1"
that's why I got Error
So I changed the region my google id to USA and change the Language than the error is gone.
I let you know the error thany you for me 🙂
I think the when we make a new sheet in
Woohoo!! Thanks for being so communicative!
Saludos
Oh is this why sharing is not working just yet?
Thanks for your feedback. I am sorry I am not sure I followed your request here. Can you please elaborate a bit? TY!
We saw a bug creep up yesterday where signed out users couldn't view opals shared with them - we fixed this late last night. So if you had any endless spinner experiences w/ shared Opals yesterday, please try again today and lmk if there's still a problem please! 🙏
Thank you Megan! I looked today and my app seems to be working
😢 I've seen several people using Opal in Brazil. Several YouTube videos. However, for me, this message appears.
Same thing here 🥲
Me too...
Is Opal not available for workspace Google account?
Heyo, do I nesd to have fear that my Google Drive data is used as training data when I connect Opal to my GDrive
Why doesn't the nano-banana image generation step support all the aspect ratios? Is there a way to make that a dynamic selection based on the input?
Help. It's been 2 days since Opal stopped working here in Brazil
Hello friends, i need help.
{
"planStep": {
"stepName": "GenerateVideo",
"modelApi": "generate_video",
"inputParameters": [
"text_instruction"
],
"systemPrompt": "",
"output": "generated_video",
"options": {
"disablePromptRewrite": false,
"modelName": "veo-3.0-generate-preview"
}
},
"execution_inputs": {
"text_instruction": {
"chunks": [
{
"mimetype": "text/plain",
"data": "<base64 encoded data>"
}
]
},
"aspect_ratio_key": {
"chunks": [
{
"mimetype": "text/plain",
"data": "<base64 encoded data>"
}
]
}
}
}
Gives Failed To Fetch . Anyone ?
I prompted "a useful translator" and it made a really useful translator 😂 I was surprised that worked
I need that on my phone lol
does anyone know for how long data generated through opal apps is kept?
Hello, we wanted to remind you of our server guidelines, stated in #start-here・📍 : “Currently, our team can support engagement in English; the discussion in this server will largely remain in English for now.” Please keep in mind that non-English discussions and content may be removed. Thank you.
Woah! Did you make a mobile browser UI for Opal?
im having trouble getting my opals that use nano banana to work, getting 503 errors rn. i tested and it was working when i made it a few days ago
I am from Arg and it says is not avalayble here help
Hi everyone! When is OPAL available in Spain?🇪🇦
im having trouble getting my opals that
data is stored for 30 days, but only used for troubleshooting and analysis (IE we will look at de-identified data to evalaute quality metrics and use cases occasionally). No training on Opal data
Is Opal not available for workspace
Help. It's been 2 days since Opal
A fix is out for all Brazil and Argentina users - Opal is back up in your countries! Sorry for the outtage, we didn't realize we had access issues for you all until you reported. THANK YOU for flagging this for us!
😢 I've seen several people using Opal
Heyo, do I nesd to have fear that my
Hi, is Opal still using Veo 3 or now it uses Veo 3.1 ?
Does our Google AI ultra subscription connect with Opal? so for example, I can generate unlimited Veo 3 fast in google flow with Ultra. If I have an ultra subscription, will I be able to generate unlimited videos through Opal? How does it work for workflows that end up in video generation? Is there a plan to integrate the subscription into Opal?
Yes, we're actually working on this integration now with the G1 team, but they're pretty backed up on integration projects. So no exact ETA yet, but the plan is to start with video - when you run out of video quota in Opal for the day, we'll check if you have AI credits or a G1 subscription so we can start pulling from those!
Why are viewers being forced to log into their Google account just use the app?
because google hosts your mini-app for you! so if you create an Opal, you don't have to put in an API key and pay for your user's AI model usage. Google gives each end user a certain amount of free access to models so you can use other people's Opals for free! In order to control that, we need to have people signed in to prevent abuse and ensure we can offer you some free AI usage 🙂
Haven't done this in a while, but posting some recent Opal feature updates! Please feel free to send over feature requests as well 🙂
🎛️ Editor controls: We’ve made the undo/redo button more prominent by pulling that out and adding zoom controls as a toolbar on your visual editor. Bonus tip: you can also find version history if you click your three dot menu > version history.
📌 Homepage Pinning: You can now pin your favorite Opals so it’s easier to find the ones you use regularly
**🔍 Homepage Search: **We moved the search bar closer to your list of Opals so it’s easier to find what you’re looking for.
**🐛 Sign-in & User Experience Fixes: **Several bug fixes and enhancements improved the sign-in process, including better error reporting, and enabling signed-out users to view Opals shared with them. More fixes to sign-in coming soon!
**🌎Brazil and Argentina access bug fixed: **Thanks to the users who reported access issues in some Latam countries - we’ve identified and fixed this bug!
When is the app creation reset?
no limit on number of apps! But if you get rate limited for a model, then it's every 24 hours
Hi everyone! When is OPAL available in
Why doesn't the nano-banana image
will these ever be deployable like the apps in the genai studio or is this just a usage experiament?
When Opal will be available in Tunisia?
lol Opal needs access to google drive, photos, everything lol
Greetings, Opal needs 2 things. Ability to restart a mini app after output and rerouter blocks, Non-Static HTML Webpages, & Ability to rerun from a particular block with a button or run entirely new blocks connected from the outputs and accessed via the output.
Also, Mobile Usage of Opal Editor Mode (Super hard)
hi
my opals prompt is ^ but its heavily still hallucinating. what can i do
the tools i use with n8n have stuff like getting frames every 0.2fps which this simply doesnt
@unique zealot That sounds frustrating. Analyzing highly edited, fast-paced videos with decoupled audio is a challenge for any AI, as the speed and lack of direct visual/audio correlation often lead to hallucination (the model generating plausible-sounding but incorrect information).Here's a breakdown of why your current process is likely failing and a potential solution approach.Understanding the ProblemThe core issue is that your tools are struggling with a mismatch in data density:High-Velocity Visuals: The video has an extremely fast cut rate. A 0.2 fps frame extraction rate means you're getting only 1 frame every 5 seconds.Calculation: $1
my opals prompt is ^ but its heavily
Anyone have any idea how to fix the opal google experiment software so it works again for photo to video?
He @soft edge
Hey @dreamy bramble
Whats good ? What are you working on lately ?
All sorts of things buddy, I run a non profit. Open Research and Development Laboratories
Ohh wow that’s awesome
I’m hoping to make a some apps I have some what I think are great ideas lol
Good, do you know how to write software or anything?
When Opal will be available in Tunisia?
Greetings, Opal needs 2 things. Ability
need veo3
need veo3
Hey everyone 👋,
Really excited to be part of Google Labs! Just wanted to mention that some tools, like Opal, aren’t available in my country (Mali) yet, so I can’t use them for now. Hopefully they’ll be accessible everywhere soon!
Thanks for understanding 🙏
Hi, for the past few days, I have been having error with Gemini 2.5 Pro or Flash Generate Card, constantly getting "The model is overloaded. Please try again later." or "Something went wrong. Please try again later.". Only sometimes, the models decided to load. Just want to see if any other people are having the same issue?
No
I wish opal could create songs like Music FX or maybe like Music AI Sandbox
have you tried using the Lyria model? You can create music with Opal!
oh really? I'll try it right now! thanks
Anyone have any idea how to fix the opal
@elfin tulip when will opal be ready for sweden??
No exact ETA yet, we're working with legal on that now!
FYI - Advanced Protections blocking is fixed now, you should be able to use Opal even if you're opted into advanced protections!
Great how long does it take push a eta (legal) usually? Why does take so long time get preview mode and feature to Sweden is it possible sign nda or try the experimental Modes early or Early access preview mode so much fun project possible to with other ai application already. Think even google pixel feature are so delayed in sweden? Why is that when some feature is available eu and uk.. sweden is in eu so should get safe or better feature hopefully?!
hi <@&1173646484311384114> any chances of knowing when Opal will be available in New Zealand?
my org has n8n workflows that use google integrations, would be interesting to see if Opal is a better and more secure alternative
hi @Google Labs team any chances of
How generate video in opal 9:16 aspect ratio, in settings change but its don't work
Hi all, do you have a roll out time for users in México?
Hi all, do you have a roll out time for
@Megan Li (Labs) when will opal be ready for Mexico? if is posible know
How to publish our opal without showing the editor tab?
Every time i login opal, it shows this message, anyone knows how to deal the problem?
Why doesn't Opal have a bug or defect feedback section in discord?
How to publish our opal without showing
Is there any mechanism to do a while loop?
For my case, I'm trying to upload bunch of documents at once, but I think it is impossible to do it in one go
Thanks in advance
Hello, I’m back and now it launched in Brazil, so I’m back full force again.
Right now, I am focused on creating a prompt refinement for Sora. I do not know if it is allowed to talk about it here.
Yes, I’m back.
@quaint eagle I missed you, my friend.
Can you share the workflow created for generating ready data using Sora?
not right now, but we're prototyping an agent step that would enable looping
@Megan Li (Labs) when will opal be ready
what are the limitations of opal? I submitted an app concept but it declined the proposal and said "this application requires tools that are not supported"
what are the limitations of opal? I
Hello, I am trying to create a social media automation, thought I am able to create the post fetching input from google sheet, I am unable to update it in the same sheet.
Basically, the app shows the output but fails to update, is there need for any sheet api to be enabled to make it work?
can anyone help me wit this?
Hello, Megan. Can this Agent Step use Interactions from the Output to do so and can we connect it to a different setup?
Like Lets say, I have an unconnected setup which does lets say, Give me Audio for a word with the example using the Output as The Input and Give the Audio?
Hi everyone, I'm new to Opal
I am trying to build a kind of "Social listener" that will browse all the internet and look for upcoming free Music Events
I believe there should have been some similar app, but I have not found it on the gallery yet
The gallery is also not very diverse, it contains only apps build by the Opal team
Is there any other available sources that I can look at and build my app? It would be great if we have a community-contributed source
so timely! Was just having this conversation with another Opal user who was asking for a community gallery. We're scoping this out now, thanks for the feedback 🙂
In the meantime, have you tried making the app you described? Definitely sounds useful! If you create an Opal, you can try with natural language writing in a bit more detail the use case you just described. Lmk if you need help once you try it! @pallid carbon
Hello, I am trying to create a social
Like Lets say, I have an unconnected
I am learning and trying! Thank you for your support! I will get back as soon as I get the first version work. So excited that I may post the very first community app on the gallery haha
When is this coming to Sweden?
How should one view opal vs openai agent Builder for vibe coders?
great question! agent builder seems more like an enterprise/dev-focused agent builder tool for companies to build agents in production. it's also more of a technical UI. Opal is meant to be easier to use for anyone to build workflows, agents, and mini-apps that solve every day problems, automate tasks, or are fun to share
opals are also hosted by google so you don't have to worry about setting up an API key or deploying your agent when it's ready to be used by others. You can share an opal with one click and google handles the rest 🙂
I'd like to know if there's a possibility of using conditional blocks? I want to automate an agency, and depending on the task, it will be routed to a different block.
Yes, You can
On The Generate Step right after the input just put a lot of If Else Statements in Natural Language call this step "Interpreting Prompt"
Failed to fetch
Hello @ruby gale , a friendly reminder, you cannot generate images or videos in this server. If you are looking for prompt help or guidance, please make sure to ask clearly and to do so in the correct channels. Thank you!
Example: the user has the option to choose whether a video or an image will be generated. For example, the user chooses the image in the initial input. However, I only want the image generation flow to be activated.
My Opal App can't be accessed publicly despite the fact that I added permissions to "Anyone with Link". Anyone knows why this happens?
Heres the Prompt for that:
"If @User_Input has the word "image" run @Image_Function
and
If @User_Input has the word "video" run @Video_Function
Else: Run @Error_Output with the error: 'Please Pick either a video or image'"
Good morning, we also want access to Opal in Romania
hello
@tough oyster As stated in #start-here・📍 , currently, our team can support engagement in English; the discussion in this server will largely remain in English for now.
yess
veo 3/3.1 gives "failed to fetch" error
Is Opal available in Paraguay?🇵🇾
My Opal App can't be accessed publicly
separately, we're coming out with native support for conditional logic with an agent step soon!
Hi team I have question : can I use images generated by Google labs on tik tok for commercial purposes or no thank u ♥️
their is a release date for the MCP ?
hi
any updates on loops?
iteration?
sometimes or most of my app ideas need to generate multiple things
Hi team I have question : can I use
no exact date yet, as we need to work through approvals on our side in addition to the feature development work!
any updates on loops?
Anyone know when / if this will release to UK? Seems like such a useful tool.
hi
any documentation for how code execution works in opal?
i got it working with python code in the generate node but would like to understand a bit more how it works to best use it
cc @elfin tulip
ps. opal is awesome! we’re running an internal workshop next week for rev ops and it’s been so easy to get started. great job 🙌
Any plans to deploy our app’s as an api endpoint? (Just like n8n)
we've thought about this, but not on deck in the short term!
Recorded a quick tutorial for anyone new to Opal! https://www.youtube.com/watch?v=NWNNDvehBIU
Megan Li - Senior PM @ Google Labs - walks through a tutorial of Opal's visual editor. Learn how to create a new AI workflow or mini-app using natural language and a visual app builder.
Subscribe to Google for Developers → https://goo.gle/developers
Speaker: Megan Li
Products Mentioned: Gemini
Is there an estimated release date?
helpful to have the full walk through in one place! thanks for sharing this
didn't know it could support multiple user inputs for multiple questions in a flow like forms.
--> wonder if there's a way to make that more intuitive, like nest additional questions in 1 user input node (like Gumloop interface node)
also any way to create new google doc as an output for each run, or will it always update the same doc?
--> would be nice to have a toggle to decide per workflow
👋 Hello everyone,
I'm new to the Opal community and very keen to learn!
I'm starting an experimental project using React and React Flow to create a workflow-based automation program, connecting modules via custom nodes and edges. I'm proficient in JavaScript.
I'm interested in finding out if there are any contributing tasks I could potentially get involved in. I'm eager to learn and grow with the community's guidance.
I am also a member of the Google Developer Program (Premium).
Hello. I am new and tried the platform today and was encouraged. In testing I could not see a way to copy/export any of the created text made by the platform. Only a voice clip am I able to download? Is there something I am missing as I cannot seem to use the creation.
#opal-general・✨ When you create content on the right side (the "Immersive" or "Canvas" editor), it's designed to be a standard text or code editor, which means you can simply copy the text directly using the standard keyboard shortcuts or right-click context menu, just like you would in a document on your computer.
Here’s how to do it:
Select the Text: Click and drag your mouse to highlight all the generated text you want to export.
Copy: Press Ctrl + C (on Windows/Linux) or Cmd + C (on Mac). Alternatively, right-click the selected text and choose "Copy."
Paste: Paste the text into any external document, email, or other application using Ctrl + V or Cmd + V.
The ability to download voice clips (audio) is a specific feature, but for the created text, code, or Markdown documents, you just use the standard copy/paste functionality from the editor window!
I cannot seem to select the image within the image tool.. How on the image?
Always reply on the same thread :
If the standard right-click doesn't work, here are two more ways to grab the image:
Use the Preview Button (If available): If there is a "Preview" or "View in Browser" option for the Immersive, the image will open in a new, uncluttered window, where the right-click "Save Image As..." action is almost guaranteed to work.
Use a Screenshot: This is the universal workaround.
Windows: Use the Snipping Tool or press Windows Key + Shift + S.
Mac: Press Cmd + Shift + 4.
Mobile/Other OS: Use your device's dedicated screenshot function.
Thanks.. the right click is not working for me on a Mac.. also no “Save image as” not available
I also get this error of video generation: Video generation expects either a single text description, or text plus a single image. Got 2 images.“ I tied a couple times
That's totally fair! Here are the quick solutions for both issues:
- Image Export (Mac)
Since right-click save is blocked, the easiest way is to use the built-in Mac screenshot tool:
Press Cmd + Shift + 4 and drag a box exactly around the image to save it instantly to your desktop.
- Video Error Fix
The error "Got 2 images" means the video tool only accepts:
Text only (to generate a video from scratch).
Text PLUS one single image (to animate that image).
Action: Go back and ensure you only have one image (or zero images) selected before you try to generate the video again.
helpful to have the full walk through in
Hi Opalites! We expanded <@&1410743137319522395> access to 160 new countries today 🥳 🌍 (EU folks, coming sson!). For those who are curious, we also published a blog post with categories of what users have been creating with Opal so far - https://blog.google/technology/google-labs/opal-expansion-160/.
For new users, bumping up the new tutorial for a walkthrough of Opal features: https://www.youtube.com/watch?v=NWNNDvehBIU
Say hi and let us know if you have questions or feedback ideas!
Megan Li - Senior PM @ Google Labs - walks through a tutorial of Opal's visual editor. Learn how to create a new AI workflow or mini-app using natural language and a visual app builder.
Subscribe to Google for Developers → https://goo.gle/developers
Speaker: Megan Li
Products Mentioned: Gemini
When eu???
Exciting stuff! Looking forward to build more on Opal
Does this include the UK?
Exciting!
When Poland?
sorry EU folks, I promise this is in the works!! just taking a bit longer and no ETA (I am asking!)
Preview can join please
no preview, but working on unlocking general EU access so everyone can benefit
It's cool to have access to Google's Experiment AI Tools. I have been waiting for this 
Can use it with my ultra sucription like flow?
For EU users you can still use a VPN to have a access to Opal 🙏🏻
How can we get it faster in Sweden???
I recommend Nord, or.... Proton If you want something that's free
Won’t the people who created Opal get mad then?
I think Opal's current situation in Europe is due to the numerous restrictions they impose on Google. And it's even worse in France, where I am, because we don't even have access to AI mode on Chrome...
is there any way to see which 160 countries?
🤔
You have to guess and @elfin tulip will tell you hot or cold and if you get one right

For bro help me out
Sweden when????
For the communications team: when a Labs feature expands to other countries, could you include a list somewhere? Because saying “Opal is now available in over 160 countries” is kind of meaningless on its own, and then you end up with questions on Discord like “Sweden when?”. Thx.
Belgium seems not be one of the 160?
I think the EU is still out for personal data protection reasons. I'm sick of it
Hi! good day everyone. Im wondering. since opal is related to mini apps. how i can connect an api? either to get data or send data. thank you.
CONTEXT: I have an automation on n8n i cant complete due the lack of time and knowledge. One thing I do like a lot is the prompt interface from Opal. What i want to do, is get from x account via the api a message. The automation will run every 15 minutes. When the account is mentioned it will then extract the post content text. Then based on the text from the post it will perform 1 action, or another action, or another action.
I will be very specific, the account is for extracting 3 types of text and then do an if in the automation based on this."extract the prompt" or "extract the prompt from the image" or "extract the prompts".
The automation should only run when a specific account is the one tagging this account. After the text is filtered if its the text of "extract the prompt from the image" its very likely the action is done by calling an API endpoint. it would be ideal i could use a native ocr system.
I dont have figure out how to. I was thinking either hosting in a vps gemma 3 and pass an image to get the text extracted or code my own ocr implementation and put it behind an api endpoint. After the text is extracted, it should be posted in firebase realtime database by using a service account. For this I can create another fastAPI powered api endpoint for just posting into the database.
I have a very complex automation as you can see, with many different sub parts that connect. But this is not possible currently. I feel now opal is just for very very very very very simple automations that rely a lot on AI. As a solo entrepreneur I need very cheap and affordable solution it can scale if the automation scales in terms of usage.
its very unlikely I will adopt Opal if relies a lot on paying models. Thanks for any feedback.
So IDK if possible but i want to be able to trigger opal automations via api endpoint and communicate the opal automations to multiple api endpoints. if is possible please advise
aditionally i found a few bugs. BTW 'm always open to give feedback as I've done before with other producs as Lydia. When I click the node, when i attempt to customize the UI gets cut.
Congratulations on the newest expansion, Opal 🎊🎉🙌 160+ countries is no small feat.
totally fair feedback, will post a list in FAQ. In the meantime, for you all here's the list (broken up into multiple message):
(1/2)
Australia
New Zealand
Chile
Kenya
Mexico
Philippines
Puerto Rico
South Africa
Taiwan
Thailand
Algeria
American Samoa
Angola
Anguilla
Antarctica
Antigua & Barbuda
Armenia
Aruba
Azerbaijan
The Bahamas
Bahrain
Bangladesh
Barbados
Belize
Benin
Bermuda
Bhutan
Bolivia
Botswana
British Indian Ocean Territory
British Virgin Islands
Brunei
Burkina Faso
Burundi
Cabo Verde
Cambodia
Cameroon
Caribbean Netherlands
Cayman Islands
Central African Republic
Chad
Christmas Island
Cocos (Keeling) Islands
Colombia
Comoros
Cook Islands
Côte d'Ivoire
Curaçao
Democratic Republic of the Congo
Djibouti
Dominica
Dominican Republic
Ecuador
Egypt
Equatorial Guinea
Eritrea
Eswatini
Ethiopia
Falkland Islands (Islas Malvinas)
Fiji
Gabon
The Gambia
Georgia
(2/2)
Ghana
Gibraltar
Grenada
Guam
Guatemala
Guernsey
Guinea
Guinea-Bissau
Guyana
Haiti
Heard Island & McDonald Islands
Iraq
Isle of Man
Israel
Jamaica
Jersey
Jordan
Kazakhstan
Kiribati
Kyrgyzstan
Kuwait
Laos
Lebanon
Lesotho
Liberia
Libya
Madagascar
Malawi
Malaysia
Maldives
Mali
Marshall Islands
Mauritania
Mauritius
Micronesia
Mongolia
Montserrat
Morocco
Mozambique
Namibia
Nauru
Nepal
New Caledonia
Nicaragua
Niger
Nigeria
Niue
Norfolk Island
Northern Mariana Islands
Oman
Palau
Palestine
Papua New Guinea
Paraguay
Peru
Pitcairn Islands
Qatar
Republic of the Congo
Rwanda
Saint Barthélemy
Saint Kitts and Nevis
Saint Lucia
Saint Pierre & Miquelon
Saint Vincent & the Grenadines
Saint Helena, Ascension & Tristan da Cunha
Samoa
São Tomé and Príncipe
Saudi Arabia
Senegal
Seychelles
Sierra Leone
Solomon Islands
Somalia
South Africa
South Georgia & the South Sandwich Islands
South Sudan
Sri Lanka
Sudan
Suriname
Tajikistan
Tanzania
Timor-Leste
Togo
Tokelau
Tonga
Trindad & Tobago
Tunisia
Turkmenistan
Turks & Caicos Islands
Tuvalu
Uganda
United Arab Emirates
U.S. Minor Outlying Islands
U.S. Virgin Islands
Uruguay
Uzbekistan
Vanuatu
Venezuela
Wallis & Futuna
Western Sahara
Yemen
Zambia
Zimbabwe
So IDK if possible but i want to be able
Can Opal be used the way that I name the stack I want to use to create the micro app?
Like Python based Langchain / Langgraph, API access via FastAPI?
I have multiple ideas but 0 coding exp and really need someone to basically hold my hand along the way. I really think some of my app ideas would do really well.
Also how do we know for sure someone’s not going to take our idea on opal and run w it and truly build it out ?
Also good afternoon 
Can Opal be used the way that I name the
I have multiple ideas but 0 coding exp
not in italy, that's unfortunate 
Hello. I created an elaborate output to test things and when I clicked from console to step, all of the data disappeared? Where is it all as I cannot locate any of my previous outputs. Are they automatically saved anywhere? Do we need to immediately copy and paste all outputs?
Opal is not working properly i gave one automatic work and it will not be creating a working
Just go for it man!
Iteration its the best place to start.
These is the tool.
These is the time bro.
These is it.
And also preview is not coming properly and format
I got the same thing but I believe I was using the ones by default and I did not see how to save those
I think the beta might have some things .... I guess its the beta, right?
Lets have some grace and try it around it.
For example on my Opal Im not allowed to copy and paste anywhere I guess its a bug right?
I´m no AI lawer but we are asking AI to be spotless, perfect, 100%. Lets not forget people are still build it ..... right? jaja
Its mind blowing what can be done
Thanks
The no copy paste thing its hard, specially since once the prompt its send, its gone jeje
Would be the possibility to integrate with different tools outside Google ecosystem?
I would belive that way,
And thinking about it just having one only with Google tools might be cool, read some admin team reply about No-Code tools and see how far can these go ....
🚀
PS: Another option where a copy paste its very usefull. Obv I just screenshot and paste but still ...
Cheers and thanks!
Lets keep it up!!
Just noticed this released to more countries...
I tried this and I don't get the use case for this?
Is this something like Gemini Gems with more functionality?
All text or prompt based progression basically?
@azure wyvern
Basically, The Ability to use AudioLM, Veo, Gemini Flash all via single prompt in formatted manner (Google Docs, Sheets, Slides, Static HTML Page, Regular Text Layout) etc.
You can also create music, Use Drawings as Inputs and Reference Files via the Assets.
Currently a little janky for mobile web version (No Editor Mode Access).
Chalkboard Brew was playing
That’s what I’ll do soon as I finish up w this annoying job I have. It’s only weekends and 2 extra days cause I’m Trang since a promotion.
Soon as I finish up I plan to spend half the day TradingView and the other half on this.
I need more info on how to create music w drawings and such.
This is so incredible! Great to have a community as well.
Hello, I signed up for Opal using my Google account. Since then, I've been consistently encountering blank white screens whenever I try to access the application.
Hello, I signed up for Opal using my
I clearly have no idea. How to do that. The Lymia Model is for Music Tho & Drawing is one of the inputs
Hooray Opal! Nice writeup
https://chromeunboxed.com/googles-no-code-ai-app-builder-opal-just-got-a-massive-global-expansion/
When Google Labs first unveiled Opal, its no-code AI mini-app builder, we were incredibly excited. It's one of those projects that truly feels like the future, giving anyone the power to build custom AI tools without writing a single line of code. After a limited initial roll-out in 15 countries, Google saw what we all...
I am currently encountering the following error message below. It mentions something about quotas, but I'm somewhat confused as I haven't used Opal since September and the opal/asset I'm using are very simple (a sheets file with very little data). Has anyone encountered this before or knows what to do about it?
Unable to handle asset drive request: { "error": { "code": 429, "message": "You exceeded your current quota, please check your plan and billing details. For more information on this error, head to: https://ai.google.dev/gemini-api/docs/rate-limits. To monitor your current usage, head to: https://ai.dev/usage?tab=rate-limit. \n* Quota exceeded for metric: generativelanguage.googleapis.com/file_storage_bytes, limit: 21474836480", "status": "RESOURCE_EXHAUSTED", "details": [ { "@type": "type.googleapis.com/google.rpc.Help", "links": [ { "description": "Learn more about Gemini API quotas", "url": "https://ai.google.dev/gemini-api/docs/rate-limits" } ] }, { "@type": "type.googleapis.com/google.rpc.QuotaFailure", "violations": [ { "quotaMetric": "generativelanguage.googleapis.com/file_storage_bytes", "quotaId": "FileStorageBytesPerProject", "quotaValue": "21474836480" } ] } ] } }
I've also run some of the "gallery" opals just now, and there don't seem to be any issues running them. Any help greatly appreciated!
Nevermind, it appears whatever caused the error has resolved itself (suddenly started working without the error despite nothing changing 😬 )
I think the beta might have some things
Just noticed this released to more
Sign in - Google Accounts
Privacy settings - my workplace has privacy locked down in our Google Workspace, but I can't tell is Opal is 'inside' our workspace and follows the stricter no-sharing/no-training that our workspace has, or if Opal is outside of Workspace and follows the more general 'ya, we may train on it' looser privacy settings.
If I have to guess, I would guess the second one.... As Im pretty sure you are treating as such and already have a strong believe it is outside your workspace rules.
Just wish I could get an official word. I'd love to be able to use Opal for more secure things.
Yeah Opal its sucha powerfull tool!
Hope you get the response you want 🤞
Privacy settings - my workplace has
Is there any way to do conditional things with opal? I want the final output to be between two options but im not sure how to do this
Hey
Has anyone successfully connected Opal to firebase?
If using Opal though VPN- have you encountered any issues connecting to different services?
Might this be the reason I encounter difficulties connecting to firebase [screenshot]-app is published, but when i click "View Share Permissions" this is what i get
so I was generating some video clips with flow and then after I ran out of credits, I just tried using opal (as I knew that it offers a few free videos). I was kinda confused with the quality of the videos made by flow (with 3.1 fast) and opal. I wonder if opal made it with 3.1 quality version (as the video quality was much more detailed).
Hey team Opal,
Bit of a question ..... I am the only one that can't paste as Crtl + V or left click in the text box?
Its becoming a bit of an impass on
having the same problem, cant paste
Hi
When Sweden? Anyone knows
hi, can I deploy my Opal model to my own website django based?
👋hi
Hey team Opal,
When will this be available in Austria?
hi opal team, i am facing an issue where the data retrieved from google drive sheet is mostly stale cache but sometimes the data is updated
hi opal team, i am facing an issue where
Just started on Opal and it seems I still cannot integrate with tools from MCP Servers.. When will this be available? (I am In India)
Also if I try to use it with my organization account, it gives error:
The "Complete Payment" goes to the Cloud identity page - no clue why it is not allowing when it works fine with my "free" gmail account!
Are UK users having troube logging in? I keep getting that it is not available in this country
yep gotta wait for our turn unfortnately
Oh well, thank you for the answer
#opal-general・✨ love
Isn't Opal supposed to be available in Spain now?
hey all! Opal's available in 160+ countries! check the full list here: https://developers.google.com/opal/faq
I've just created a thread to keep track of region requests! please share in this thread which places you'd like to see Opal expand to ✨
Regions Thread: Opal
hey, is it normal for Opal to lag a lot with a lot of nodes and links between nodes? it's barely usable for me when i was trying to make a complex workflow
hey, is it normal for Opal to lag a lot
Also if I try to use it with my
hi
Found a significant issue. I duplicated an app, then modified the cloned one for a different but similar purpose. I saw in the dashboard that both apps were there. Next time I visited the dashboard the original is gone even after several refreshes. I 100% did not delete
Found a significant issue. I duplicated
Hey! Im new here. How can i check out what you’re building? Excited to test out
Opal sometime just doesn’t load
Hey! Im new here. How can i check out
I gat issues to report
Welcome in! So happy you’re here!
Shout out to the Opal team, had fun sharing what I’ve been working on at the Behavioral Health Tech conference this week. Get ready for more friends in the discord @elfin tulip 💃🏾💃🏾💃🏾
Hey everyone quick question on video generation in Opal: is it possible to generate videos which are longer than eight seconds I want to create videos which are like 45 to 60 seconds thanks
That is not possible by Veo
Veo only supports 8 seconds videos now
Opal directly calls Veo via API so the video length depends on Veo
Hi! Tried the City Builder tutorial but get error "Calling tools generated the following errors: "Cannot read properties of undefined (reading 'data')". Any ideas? tia!
Is Opal open source? if yes I want to apply Cryptography for modulo in google opal, you can test again, this has passed the stress test and 51% attack but again for anyone can import or implement this, named SAI-288 or sai288, I hope the relevant team evaluates first or considers this if it is possible.
we need opal in turkey plss
Hi All,
to the admins or the developers of Opal, I have a question.
Is there a way to get a local instance of Opal, like download it locally and work on it. So that we can see the source code of apps being built?
not just devs...if anyone who knows this also, please let me know
Gems + Opal = match made in heaven. Great for stock trading.
i want to create the "Ingredients to Video" like feature in my Opal Video generator app. Is it supported by Opal to do that?
Hey guys I wanted to ask something, if i make a mini app with opal, can I download it to my phone or I can't? Thank you and have a great day
Hello~
This is a question that arose from taking sales information from Google spreadsheet and performing monthly comparisons.
Is there a way to make a workflow or created app run at a specific time, schedule setting?
Overall, I was very satisfied to see that they used no-coding to design, design, and even create an app.
However, is there a way for this workflow to run (results) faster?
Every time I run it using this organized workflow, the design changes, like insights, is there a way to fix it and print it out?
You can run it via your web browser from your phone. The interface, for the editor, is designed for desktop and is very challenging on a phone.
Does anyone have a good opal for imagen generation and video in a single run?
Great questions! and cool use case 🙂
We're working on consistency, so that should come soon for the output layout. We are also working on making things faster and adding triggers so you can automate Opal runs, but that's likely coming a bit later, so closer to the christmas or the new year
Shout out to the Opal team, had fun
Does anyone have a good opal for imagen
Suggestion: Allow for MCP Server inputs to set off a flow. This would essentially allow Opal to be agentic which would allow for automations from within any application
Update: or even allow for an api call to trigger with the Opal flow id along with any inputs.
I guess you'd also want to be able to connect to a MCP server on the output too so users could leverage opal flows and send the response wherever
opal overload too much
Opal often runs into overload issues. I hope the team can fix this, as it happens too frequently, especially during analysis or script creation
ota
Hello, do you know when Opal will be available in Romania/Europe?
ai studio build is completely unusable today. does not apply fixes. breaks upon first response, cannot follow basic commands. feedback button doesnt submit. been using it for 3 months this is messed up. what did you guys do?
same, I was facing ai not being able to fix errors
Just wanted to report problem with Opal not sending the completed task to my Google Docs. It did the first time but not the second time.
It adds it to the bottom of the same Google Doc! Working on changing this behavior so it's more clear where the output went
ai studio build is completely unusable
thanks for the feedback, we are working on getting more model capacity! Can you share which models you're seeing the most issues with?
Suggestion: Allow for MCP Server inputs
Can I use gemini 3pro here to build the frontend and conversations? 3pro's conversational capabilities are much better than 2.5p.
I want to create a comparison app that uses Opal to distinguish the capabilities of two chat models.
hi
getting this a lot "Calling tools generated the following errors: "Cannot read properties of undefined (reading 'data')""
Anyone tried the antigravity ide yet?
getting this a lot "Calling tools
Located in Norway. Opal supposed to be available in 160 countries 2 weeks ago? Surprising if Norway is not one of them? 🤯
Located in Norway. Opal supposed to be
I built a successful project in Opal, but every time I try to publish/share it spins for hours and doesn't publish. I tried copying the project (remix) but have the same results. I granted full permissions to GDrive per instructions. How can I publish and share the Opal project?
please open a thread in: #1238621511401734254 for more help!
When Eu
If you haven't already please make your country requests here: https://discord.com/channels/1172568727942860810/1438243860454182922
Opal never finishes loading the landing page in Chrome on iOS for me today.
Repeated attempts over several hours. Rebooted phone, restarted chrome, cleared all tabs and made fresh one. Same experience on iPhone & iPad.
Opal worked from my PC in Chrome on Windows 11 earlier today.
Is this a known issue?
Hey guys
Once my app finalized it's deep research report, I can't scroll through the report often. Is this just an app overload due to traffic issue?
gmgm
Same thing none of my opals are loading when I shar them I wanted to do so today and the people who got this links could not load it looks to be just the basic opal projects load only ATM.
Is opal available to Indian users now?
Yes
Been Available for Indian users since a few days before Diwali
When I log in to opal.google on iPad Pro Safari and Chrome, it just keeps showing loading.
iPad Pro 13 + iPadOS 26.1(23B85) + Chrome version 143.0.749.38 + Safari
Please add support for Gemini 3.0 and Nano Banana 2, begging you!
It would be nice to have a Decision Point to choose between one or more different routes/outputs instead of relying on the output step to distinguish that
Please add support for Gemini 3.0 and
It would be nice to have a Decision
where is the best source for me to learn how to maximize using opal / vibe coding with gemini?
imho, learning about software architecture, design of UX would be great. Also knowledge about variety of algorithms and their property called complexity would be also valuable. That would help you direct the thing into the right direction, and spot eventually the issues with performance to also tell them (AI Agents) to review and fix. For example, I know some Flutter and know about rebuilding widget trees. I specifically told about this to remind the Agent about those property of flutter. Also, there is many many nice sites and books about refactoring. Which is basically the thing does after the building software, and when you say "make the code simpler to maintain", or "refactor to make performance better" - which means it should hold the same functionality but fight with tech debt. If You have any questions would be great for me to answer them @heady plank
last days, I made some apps with Vue and also used other AI Agents to review the code Jules or any other (no ads here? 😉 ) I have, created.
wait, Opal available in 160 countries, but not the UK? so many labs experiements are not available here, and so many I want to get hands on with, as many could help directly or indirectly with challenges with ADHD. sigh
Jersey, Guernsey and Isle of Man are on the list. Damn... What gives? I noticed many European countries are not on the list. First thought was GDPR causing issues, but those 3 UK Governed territories being on there make no sense...
please request a country here: https://discord.com/channels/1172568727942860810/1438243860454182922
I know, I have done for Stax for example a long time ago. It'll happen when it happens, given that it seems Europe is the intended omission. I was more asking for clarity on why many labs solutions are not available in Europe.
could you also add Italy among the Opal countries please
please request a country here: https://discord.com/channels/1172568727942860810/1438243860454182922
Opal is not available everywhere? Coz I've noticed lots of errors lately or yeah it's available here
May i request "strongly typed User Input" to allow two opals to "chain together" such that the first opal's structured output can be used for the second opal's structured input?
e.g. opal one creates a "Holiday shopping list" and opal two can "add a 'holiday shopping list' list item". Helpful for more complex data types that may have nested data
May i request a "Container" to group sub-tasks together that are truly co-reliant. Visually a user may just see the container until they "expand" it to show the sub-steps
Similarly, SSIS has both "Data flow task" and "control flow tasks" that help encapsulate and separate distinct layers of ETL logic.
I agree with this question. If i were in a Gemini chat window or an AI Studio build session, how can i easily "Call" a pre-defined opal without challenge to help integrate with the final product
On this note, it would be nice if the "input" in general could be a simple "choice" dropdown selection so it forces the users not to type in a prompt but select from a list of pre-defined options(or combobox).
May i request "strongly typed User Input
Hi!
I'm hitting a wall trying to upload large files (GBs). It feels like the upload method is restricted to a few MBs (like a Base64 payload), not supporting larger files via a dedicated system. Are you guys using a dedicated "Files API" for uploads
By the way, could you please consider increasing or enabling larger file uploads in Opal? Right now the 20MB limit is quite restrictive. I often need to analyze long videos (100–200MB), so higher upload limits would be extremely helpful for real-world workflows.
Thanks again!
I have a question, when can Gemini3 be added to opal?
Also nano banana pro
Anyone know the difference from tool: search web an get webapage?
Generate videos using the veo3.1 model using text
Generate videos using the veo3.1 model using text
regular nano banana is already there i believe, been having some fun with it, will be nice to have pro
from my playing around, the "get webpage" is more for for getting the full content. Likely to manipulate the html/styling or perhaps to merge a combination of pages into one.
The search webpage is not going to return html but instead return natural language context from that page based upon the prompt.
Yes, Ik, I was referring to the pro version
both Gemini 3 and Nano Banana Pro are coming in next week's release (Wednesday)
Thanks, Dimitri. It would also be really helpful if you could add the ability to run queries in Google Ads. For example, I'm creating a workflow for my blog and I want to make it really effective, including for SEO.
So, at a certain point in the workflow, I'd like to configure the agent to make a query on Google Ads based on the topic and language, searching for relevant, related, and long-tail keywords. Once extracted, these keywords are passed to the writer and so included in the article. 🙂 That would be a game-changing feature.
Can someone help me with OPAL and best way to use it for animated videos
"Hello my friends, I want to ask you something: does Google have an AI that makes a video from an image? Please, if it exists, send me the link or tell me its name."
Hi, I'm trying to build a small app with Opal to generate images. It checks the local time of a given country on the internet, and if it's daytime it should generate a bright image, and if it's nighttime it should generate a dark one. I’ve completed all the steps except for the initial decision block. I’ve described it to the AI multiple times, but it keeps running both branches instead of choosing only one.
Do you have any idea how I can design this decision block? Thank you in advance.
@quaint eagle According to Dimitri, expecting next week.
I had the same question: without Nano Banana Pro and the image generation is pointless.
"Hello my friends, I want to ask you
Hi all! Does any one when Opal will be available in Greece ?
I am Executiv level AI Aducator I would realy like to work with it in order to see the capabilites and see if i can propose it to my clients.
any related information will valuable.
Thank you all in advances
Hi! Will it be released in Romania soon? Thank you!
Pls add option to hide preview on the right when we are in editor view
Pls add deep research with gemini 2.5 pro and/or gemini 3 pro
Hi @𝙐𝙩𝙖𝙧𝙞𝘿 i think the feature request for a decision point is exactly the solution you need to run either a "dark" branch or a "light" branch.
Pls add option to hide preview on the
Can someone help me with OPAL and best
Hi! Will it be released in Romania soon
Hello, I am experiencing a persistent login timeout with Opal. I am unable to access the application after entering my credentials. I have already tried clearing my browser cache and using Incognito mode, but the issue persists. Could you please advise or confirm if there is a known server issue?
This issue is exactly the same thing I’m seeing
Hello, I am experiencing a persistent
Hello! Why I cannot find any user's apps in Opal gallery, as it was possible before? Is there any changes in Google politics about user's apps publication in Opal? 🤨
Add NotebookLM to that list
Hello! Why I cannot find any user's apps
working on it! also notebookLM request noted 🫡
Hey opal team, just wanted to let you know the business profiler app listed on the home page does not exist. It sends the user to the book recommender. Also, If you select an app and then click the 'View more Opal apps' link, you get an error page.
When open access in EU?
No videos generated. Opal has limited quota and it was exceeded. Please try again later.
Is generating one 8-second video already the maximum daily limit for Opal?
When open access in EU?
When is TS coming to Croatia?
Can we have 5 new outputs? PDF, Markdown, Google Forms, Colab Notes, Jamboard (Decent for Mind Maps and Whiteboarding)
And Opal should be able to edit Google Forms Code like how we do it in Apps Script
- Access to Google Photos
I got so a few QoL Ideas as well such as Group (Behavior Similar to Named Ranges in Sheets with an Protect Feature so that the AI Doesnt edit things we dont want it to edit)(Thats Means The Looper Logic can be more simple and The Group Name will pop up on the loading screen instead of names of every step and you can set up buttons like NotebookLM to pretty much enter an Automatic-Semi Automatic Prompt to run a certain Group. Another Solution is doing the buttons near the input text box. (The Reason for Buttons is for Reducing the Need for Routing Which Action to do)
We need OPAL in Spain
May i request a feature to include a "Github Repo" in the list of input advanced options? Mostly just to get a "visual icon" to see the different input type.
feature request? Any possibility of an Output Type being Jules? See my show/tell Opal as an example.
e.g. Opal does some work and one output step is input to an agent/Jule.
Can anyone suggest some interesting use cases for how to use Opal as web developer?
Hi do we have NBP on Opal? Would love to integrate it in my opal apps/prototypes 🙂
Also I wish we could make our apps with the "Ingredients to Video" feature like thing for Veo. It's there on both Vids and flow now. Also gemini app also supports something similar. So would be really cool if the same could be done with Opal!
hey @ivory patio, take a look at the gem i created in the show-n-tell to help with software development and Agentic AI
City Builder keeps failing after the first two input prompts (Location and Genre)
i just had the same issue trying to call the "Search google maps" action
Oh, that looks pretty cool. Thank you
Oh, ok. So, it's a bug then?

ooh, that's a bug, let me track this down
Fix the ux design issue when output is genereted and if you reload the page there is no way to see again the output (at least I dont see)
Fix the ux design issue when output is
integration with google forms
I did the same suggestion, ZIOCAN (I was planning to make an Map Study tool with Opal). What were you planning to do with it?
My opal has been in this state since last week and I can't log in. Is there any expert who can help me?🥹
My opal has been in this state since
my opal login has also been doing the loading state as well
May i request a feature to include a "
Hi, I noticed that Term, Privacy and FAQ links on the Opal Website seems broken. Could anyone please verify?
Unable to use Opal on my iOS devices for 2 weeks.
Every time I try to login to Opal, I get the following loading screen that doesn’t progress even after 15 minutes. I have the same experience on my iPad and my iPhone 13 Pro.
Chrome version: 142.0.7444.148 (Official Build) stable (64-bit)
iOS 18.6.2 (22G100)
It works fine from Chrome on my Windows 11 PCs.
This loading screen occurs after I confirm Opal’s access to my Google account.
What happens when you share (not Publish) an Opal app that outputs to Google docs?
Will it create a new doc for them or update my doc?
I want to share my menu planning app with several friends but don’t want to end up with their menus/recipes in my docs.
Ahh thanks for catching that! We'll fix this
I have been experiencing this issue - I suspect it is an authentication issue with gmail
