#how to generate a video that extends off an image

1 messages · Page 1 of 1 (latest)

frozen pike
#

how to generate a video thay extends off an image

distant skyBOT
#

Hi @frozen pike. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:

#

You can do this today with the Sora 2 Video API by using your image as the first frame of the video.

1. Prepare your image

  • Make sure the image resolution matches the video size you want to generate (e.g. 1280x720 or 720x1280).
  • Supported formats: image/jpeg, image/png, image/webp.[Use image refs]

2. Call POST /videos with input_reference

In the Video API, pass the image file via the input_reference parameter. The model will lock onto that image as frame 1 and then follow your prompt to “extend” the scene over time.[Videos create][Use image refs]

cURL example

curl -X POST "<https://api.openai.com/v1/videos"> \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "Content-Type: multipart/form-data" \
  -F prompt="She turns around and smiles, then slowly walks out of the frame." \
  -F model="sora-2" \
  -F size="1280x720" \
  -F seconds="8" \
  -F input_reference="@your_image_1280x720.jpeg;type=image/jpeg"

Here:

  • input_reference = your starting image (the first frame).
  • prompt = what happens after that image (how the video should continue).
  • seconds = how long the clip should be (e.g. 4–8 seconds by default).[Videos create]

3. Poll for completion and download

  • Use GET /videos/{video_id} until status is completed.
  • Then download with GET /videos/{video_id}/content.[Generate a video]
#

This workflow “extends off” your still image into a moving shot while preserving its composition and style.

Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:

#

@frozen pike Kapa.ai is still learning and improving, please let me know how I did by reacting below.

frozen pike
#

can i do this on Android fone @distant sky

distant skyBOT
#

Let me check the knowledge base for you again. One moment! :books:

#

You can’t fully do this in the Sora Android app yet, but you can do it from Android using the API.

Sora app on Android

  • The Sora app is currently iOS + web first; one help article explicitly says “The Sora app is not available on Android.”[Getting started]
  • A later release note announces an Android Sora app, but also says some features like “creating characters, stitching, and the ability to purchase additional videos” are not supported at launch, and doesn’t mention image→video specifically.[Sora release notes]
  • Separate Sora help content states that uploading images with depictions of real people is blocked; using someone’s likeness requires Cameos.[Creating videos]

The docs do not clearly confirm that “extend off an image” (image-to-video) is supported in the Android app UI today, so I can’t say that it is.

Using the API from Android

You can do what you want on an Android phone by calling the Video API from your own app or from a script:

  • Use POST /videos with input_reference to provide the still image as the first frame and a text prompt to describe how the video should continue.[Videos create; Use image refs]
  • Then poll GET /videos/{video_id} and download the result via GET /videos/{video_id}/content.[Generate video; Retrieve results]
#

So:

  • Android Sora app UI: not clearly documented to support your exact image-extension workflow yet.
  • Android device + API: yes, fully supported as long as your image follows the rules (no real people’s faces, etc.).

Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: