Video modelsAdvanced usersUpdated
POST/v1/videos/generations
Gemini Omni 1.2 Flash
Request ID gemini-omni-1.2-flash
Protocols
POST /v1/videos/generations· Generate video exampleGET /v1/videos/tasks/{id}· Check video task exampleGET /v1/videos/tasks/{id}/content· Download video exampleDELETE /v1/videos/tasks/{id}· Delete a finished video task examplePOST /v1/videos/assets· Upload an asset exampleDELETE /v1/videos/assets/{id}· Delete an asset example
Authentication
Requests require an API key from API key management. For JSON requests, set Content-Type: application/json.
Authorization: Bearer YOUR_API_KEY
Request parameters
POST /v1/videos/generations
| Field | Type | Required | Description |
|---|---|---|---|
model | string | Required | gemini-omni-1.2-flash. Use the model ID shown on this page. |
prompt | string | Conditional | Describe the scene, action, and camera movement. When a text prompt is required, use this field or a content item of type text. |
content | array<object> | Conditional | List of reference assets and optional text items. Omit for text-only generation using prompt. Use first/last frames, up to ten reference images, or one MP4/MOV video up to 15 seconds and 64 MB. Submit these input types separately. |
content[].type | string | Required | Accepted values are text, image_url, input_image, video_url, input_video. |
content[].text | string | Conditional | Provide the prompt when the content item has type: "text". |
content[].role | string | Conditional | Use first_frame and last_frame. A last frame requires a first frame. Use reference_image for a general reference image. Use reference_video for video inputs. |
content[].asset_id | string | Conditional | Upload to POST /v1/videos/assets and use the returned id. Send kind (image, video) before file in the multipart form. |
resolution | string | Optional | Accepted values are 360p, 720p, 1080p, 4k. Default: 720p. |
ratio | string | Optional | Output aspect ratio. Also accepts the alias aspect_ratio. Accepted values are 16:9, 9:16. Default: 16:9. |
duration | integer | Optional | Output length in seconds. Accepts whole seconds from 3 to 15. Text and image references default to 10 seconds. Frame mode and extension accept 3–10 seconds. Omit for editing. In extension mode this specifies the added duration. |
generation_mode | string | Optional | Accepted values are auto, t2v, i2v, reference, edit, extend. Defaults to auto, which selects a mode from the inputs. An explicit mode must match the inputs. |
source_task_id | string | Conditional | ID of your own completed and settled task for the same model. Use either this field or an uploaded source video. |
Request examples
Set your API key in the local ICODEEASY_API_KEY environment variable and choose an endpoint that accepts direct HTTP calls. For endpoints that require a specific client, follow that client’s setup guide.
Replace example media URLs and asset IDs with your own valid assets.
Video Generations
bash
curl -sS 'https://api.icodeeasy.cc/v1/videos/generations' \
-H "Authorization: Bearer $ICODEEASY_API_KEY" \
-H 'Content-Type: application/json' \
--data '{
"model": "gemini-omni-1.2-flash",
"prompt": "Slow camera push toward a paper lantern swaying gently by a lake.",
"resolution": "720p",
"ratio": "16:9",
"duration": 5
}'
Check video task
bash
curl -sS 'https://api.icodeeasy.cc/v1/videos/tasks/YOUR_TASK_ID' \
-H "Authorization: Bearer $ICODEEASY_API_KEY"
Download video
bash
curl -sS 'https://api.icodeeasy.cc/v1/videos/tasks/YOUR_TASK_ID/content' \
-H "Authorization: Bearer $ICODEEASY_API_KEY" \
-L \
-o result.mp4
Delete a finished video task
bash
curl -sS 'https://api.icodeeasy.cc/v1/videos/tasks/YOUR_TASK_ID' \
-H "Authorization: Bearer $ICODEEASY_API_KEY" \
-i \
-X DELETE
Upload an asset
bash
curl -sS 'https://api.icodeeasy.cc/v1/videos/assets' \
-H "Authorization: Bearer $ICODEEASY_API_KEY" \
-F 'kind=image' \
-F '[email protected];type=image/png'
Upload an asset · Upload a video
bash
curl -sS 'https://api.icodeeasy.cc/v1/videos/assets' \
-H "Authorization: Bearer $ICODEEASY_API_KEY" \
-F 'kind=video' \
-F '[email protected];type=video/mp4'
Delete an asset
bash
curl -sS 'https://api.icodeeasy.cc/v1/videos/assets/YOUR_ASSET_ID' \
-H "Authorization: Bearer $ICODEEASY_API_KEY" \
-i \
-X DELETE
Responses and tasks
HTTP 202 · Task accepted
json
{
"id": "vid_1785398400000000000",
"object": "video.generation.job",
"model": "gemini-omni-1.2-flash",
"created_at": "2026-07-30T08:00:00Z",
"status_url": "/v1/videos/tasks/vid_1785398400000000000",
"content_url": "/v1/videos/tasks/vid_1785398400000000000/content",
"status": "queued",
"updated_at": "2026-07-30T08:00:00Z",
"task_id": "vid_1785398400000000000"
}
HTTP 200 · Completed task lookup
json
{
"id": "vid_1785398400000000000",
"object": "video.generation.job",
"model": "gemini-omni-1.2-flash",
"created_at": "2026-07-30T08:00:00Z",
"status_url": "/v1/videos/tasks/vid_1785398400000000000",
"content_url": "/v1/videos/tasks/vid_1785398400000000000/content",
"status": "succeeded",
"updated_at": "2026-07-30T08:01:00Z",
"url": "/v1/videos/tasks/vid_1785398400000000000/content"
}
Save id. With the same API key, send GET requests to status_url until status is succeeded or failed. On success, download from content_url and follow HTTP redirects. Both paths are relative to the API base URL.
| Field | Type | Description |
|---|---|---|
id / task_id | string | Task ID. In the creation response, task_id equals id. |
status | string | Task state, one of queued, running, succeeded, or failed. |
progress | integer | Generation progress. Omitted when the service has no progress value. |
error.code | string | Error classification for failed tasks. |
Common errors
| HTTP | Error type | Description |
|---|---|---|
| 400 | invalid_request | Check the model, parameter combination, and reference media. |
| 401 | unauthorized | Check whether the API key is missing, invalid, or disabled. |
| 402 | insufficient_balance | The balance is insufficient to submit the task. |
| 404 | not_found | If polling returns 404, keep the task ID and check the polling URL and API key for the account that created the task. It may already have finished. Do not resubmit on this basis. |
| 409 | video_not_ready | The video is still being generated. Continue polling. |
| 503 | submission_unknown | The submission outcome is unknown. Contact support to check whether a task was created before sending a new request. |
Full integration guide · See the full API guide for request fields, task states, and response formats.
Capabilities and limits
text_to_video· Supportedimage_to_video· Supportedfirst_frame· Supportedlast_frame· Supportedreference_images· Supportedvideo_input· Supportedaudio· Not published- Limits Resolutions: 360p, 720p, 1080p, 4k · Aspect ratios: 16:9, 9:16 · Durations: 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 · Reference images: up to 10 · Input: Use 3–10 seconds with first and last frames or extension. Edit or extend one video up to 15 seconds and 64 MB. Editing follows the source duration.
Model Capability Catalog · View pricing
Related guides
Video Generation API
Create video tasks from text, first and last frames, or references; poll status, download results, and delete finished tasks when needed.
Model Capability Catalog
Compare public text, image, and video models by normalized request ID, compatible protocols, supported clients, tool use, media capabilities, limits, and review date.