#What happened to image generation models?

3 messages · Page 1 of 1 (latest)

whole hemlock
#

Week ago I saw broken something what worked for months. At Seedream 4.5, I provided two images and asked to merge via OR chatroom, images were ignored, I got charged and it produce something random because images were missing.

Since, I try everyday, got charged and report request with zero answer.

Until today it worked with one image. But today not just Seedream 4.5 but also Nane banana 2 via chatrooms and one image said,

No endpoints found that support tool use. To learn more about provider routing, visit: https://openrouter.ai/docs/guides/routing/provider-selection

I would appreciate some transparency about what is going on. It costed me money and I wouod consider fair if someone said, that image editing was discontinued at chatroom or something.

OpenRouter Documentation

Route AI model requests across multiple providers intelligently. Learn how to optimize for cost, performance, and reliability with OpenRouter's provider routing.

spiral gullBOT
#

I'm an AI assistant - my answers may be incorrect and I cannot see screenshots. Please verify information with the OpenRouter docs.

Thanks for the detailed report—I'm sorry about your experience and the lack of communication.

Here's what I can confirm from the docs:

  • Models like bytedance-seed/seedream-4.5 and nano-banana-2 (google/gemini-3.1-flash-image-preview) are listed as supporting image generation and multi-image composition. The official docs say:
    "The model’s multi-image composition capabilities have been significantly strengthened..." for Seedream 4.5, and show continued support for image input and output Seedream 4.5 Nano Banana 2.

  • OpenRouter does support image input via API, using either direct URLs or base64 data, and docs provide instructions for sending multiple images in the API (not the chatroom UI) Image Inputs Guide.

  • If you saw the error:

    No endpoints found that support tool use.
    This means no routing provider currently supports your request parameters or model features for that chatroom. This can be due to provider changes, model updates, feature gating, or temporary unavailability.
    The docs clarify:
    "When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it." If no endpoints are available, the request fails Provider Routing.

What this means for you:

  • There is no documentation stating that image editing or multi-image input is discontinued, but provider support c...
#

...an change at any time, resulting in these errors.

  • The chatroom UI may lag behind API support or have stricter provider requirements, which can cause images to be ignored or requests to fail, even if the API docs say it's supported.
  • Charges may still apply if the request is processed but fails to deliver expected results.

What you should do:

  • For precise model status, always check the OpenRouter Models page for availability and capabilities.
  • For billing disputes, you must email [email protected]. Note: Billing support is only available for accounts with $10+ spend; otherwise, you'll be redirected back to Discord.
  • For persistent technical issues, please provide:
    • The exact error message
    • The specific model identifier you used
    • A sample of your prompt/code (if via API)
      This helps mods and support escalate or diagnose.

Transparency:
There's no public statement about discontinuation of image editing or multi-image support in the chatroom, but provider routing and feature availability can change without notice. For current status, check:

If you want this escalated or have further questions, please provide the details above or contact mods on Discord.

-# This is an automated response powered by Answer HQ. Use !mute to stop the bot from responding, or !close when your question is resolved.