Gemini Omni Flash
Gemini Omni Flash Video Generation
Generate videos with the legacy Gemini Omni Flash channel using text or up to 3 reference images
POST
- Use
gemini-omni-flash(or the channel’s configuredgemini_omni_flashalias) in themodelfield - Supports text-to-video and reference-image-to-video
- Supports 4 / 6 / 10 second durations
- Supports
720pand1080p;1080pis landscape-only - Async task API: submit a task, then query by task ID
This page documents the non-official legacy/FlowReverse-style
omni_flash channel. It uses the channel’s configured upstream endpoint and API key, not the Vertex AI Interactions API. Other legacy channel implementations may have different limits. The separate gemini-omni-flash-preview-official page documents the Vertex Service Account/OAuth channel.Authorizations
string
required
Use Bearer Token authentication:
Body
string
default:"gemini-omni-flash"
required
Model name. Use
gemini-omni-flash for the legacy channel. Some channel configurations expose the equivalent gemini_omni_flash alias.string
required
Text prompt for video generation.
integer
default:"6"
Video duration. Supported values are
4, 6, or 10 seconds.string
default:"16:9"
Video aspect ratio:
16:9landscape9:16portrait
string
default:"720p"
Video resolution:
720p or 1080p. The legacy channel uses a landscape output for 1080p.string[]
Optional public reference image URL array with up to
3 items. Omit it for text-to-video. Base64/data URIs and video inputs are not supported by this channel.Response
string
Task ID for status polling.
string
Object type, usually
generation.task.string
Model name used for the request.
string
Task status:
queued, in_progress, completed, or failed.integer
Task creation timestamp.
string
Generated video URL when the task completes.