Gemini Omni Flash video edit: 5 reference images vs Sume
Runway's Gemini Omni Flash video-to-video edit accepts up to 5 reference images. On Sume, a video_url edit cannot be combined with any reference field.
Runway's changelog says its Gemini Omni Flash video-to-video edit accepts up to 5 reference images to guide the edit. Sume's gemini-omni-flash-1.1 edit does not: video_url is the edit source, and it cannot be combined with image_url, end_image_url or reference_*_urls.
Runway facts are from its API changelog entry for gemini_omni_flash; Sume facts from the Video Router docs, read 2026-10-01.
What does Runway allow in an edit?
The July 1st, 2026 entry says video-to-video edits an input video of up to 10 seconds, with output matching the input's duration and orientation, and accepts up to 5 reference images to guide the edit.
What does Sume's edit accept?
An edit is a Video Router field: the model id is gemini-omni-flash-1.1, the prompt describes the edit, video_url is the source clip, and resolution is optional (default 720p). aspect_ratio is a 400 on an edit, and the OpenRouter-shaped /v1/videos body does not carry video_url at all.
| Mode | Send | Reference images |
|---|---|---|
text_to_video | prompt | none |
image_to_video | image_url (+ end_image_url) | frames only |
reference_to_video | reference_image_urls (up to 10) | yes |
video_to_video (edit) | video_url | not combinable |
Can I still use reference images with a video?
Not in the same request as an edit. The docs say that when a clip should condition a fresh generation instead, use reference_video_urls in reference-to-video mode, which takes up to 10 reference images and up to 3 clips of at most 3 seconds each. That generates new footage, it does not edit your clip in place.
What should I do?
If your edit needs a guiding image, describe the change in the prompt, or switch to reference-to-video and accept that the output is a new generation. See Gemini Omni video edit prompts for wording.
Sources
Related posts
More in Developers
- Gemini Omni store=false sync video vs Sume's 30-second sync cap
Gemini suggests background=false, store=false, stream=false for synchronous video. Sume sync mode waits at most 30 seconds, then you poll status_url.
- Gemini Omni task values mapped to Sume request fields
Google's Omni task field takes text_to_video, image_to_video, reference_to_video, edit or extend. On Sume the fields you send pick the mode; extend is absent.
- Gemini TTS caps 200 voices per project: picking a voice on Sume
Gemini TTS stores at most 200 voices per project for a year. On Sume you list avatars and use one whose voice.status is ready, with no stored-voice quota.
- Text to speech mulaw 8000 Hz: Gemini 3.8 and Sume TTS
Sume TTS can return pcm_mulaw or pcm_alaw at 8000 Hz in wav or raw containers. Gemini 3.8 TTS does it with audio/mulaw and audio/alaw mime types.
Written by Sume