venice-video
Generate and transcribe videos via Venice. Covers the async /video/quote + /video/queue + /video/retrieve + /video/complete loop, text-to-video, image-to-video, video-to-video (upscale), audio input, reference images, reference video and reference audio (R2V), scene and element support, plus /video/transcriptions for YouTube URLs.
veniceai · AI review on file
Installation
Read the source and dependencies before running a command. Copying does not install or execute anything.
npx skills add veniceai/skills --skill venice-videoAbout this skill
name: venice-video description: Generate and transcribe videos via Venice. Covers the async /video/quote + /video/queue + /video/retrieve + /video/complete loop, text-to-video, image-to-video, video-to-video (upscale), audio input, reference images, reference video and reference audio (R2V), scene and element support, plus /video/transcriptions for YouTube URLs.
Venice Video
Video is asynchronous — like audio music. Five endpoints:
| Endpoint | Purpose |
|---|---|
POST /video/quote | Price in USD (no charge, no job). |
POST /video/queue | Enqueue generation. Returns queue_id, charges (reserves) funds. |
POST /video/retrieve | Poll status or download video/mp4. |
POST /video/complete | Finalize & delete media from Venice storage. |
POST /video/transcriptions | Sync: transcribe a YouTube URL's audio. |
Use when
- You need text-to-video, image-to-video, video upscale, video-with-audio, or video transcription.
- You can tolerate async execution (single-digit seconds to several minutes depending on model, duration, and queue depth — inspect
average_execution_timeandexecution_durationon/video/retrievefor your job's live estimate). - You want to price a job precisely before committing (
/video/quote).
Lifecycle — generation
1. Price with /video/quote
curl https://api.venice.ai/api/v1/video/quote \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "wan-2-7-text-to-video",
"duration": "5s",
"aspect_ratio": "16:9",
"resolution": "720p",
"audio": true
}'
Response: {"quote": 0.35} USD.
/video/quote requires model and duration. It also takes resolution
(required for models priced by duration × resolution × rate), upscale_factor
and video_url for upscale models (video_url lets Venice auto-detect the
source duration), and reference_video_total_duration for reference-to-video
models — the aggregate seconds of every reference video you intend to send, up
to 45. Quote a reference-video job without it and you get the no-reference
baseline price.
2. Submit with /video/queue
curl https://api.venice.ai/api/v1/video/queue \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "wan-2-7-text-to-video",
"prompt": "Commerce being conducted in the city of Venice, Italy.",
"negative_prompt": "low resolution, worst quality, defects",
"duration": "5s",
"aspect_ratio": "16:9",
"resolution": "720p",
"audio": true
}'
Response: { "model": "...", "queue_id": "uuid", "download_url": "https://..." }.
download_urlonly appears for VPS-backed models. When present, the retrieve endpoint returns JSON status only — fetch this URL to download. Valid 24 h.
3. Poll with /video/retrieve
curl https://api.venice.ai/api/v1/video/retrieve \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"...","queue_id":"..."}' \
--output out.mp4
- Processing: JSON
{"status":"PROCESSING","average_execution_time":145000,"execution_duration":53200}(ms). - Completed (non-VPS): binary
video/mp4body. - Completed (VPS-backed):
{"status":"COMPLETED", ...}— fetch thedownload_urlfrom the queue response. delete_media_on_completion: trueauto-deletes after successful retrieve.
4. Finalize with /video/complete
curl https://api.venice.ai/api/v1/video/complete \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"...","queue_id":"..."}'
QueueVideoRequest fields
Availability depends on the model — check GET /models?type=video.
| Field | Type | Notes |
|---|---|---|
model | string | Required. |
prompt | string, ≤ 2500–3500 | Required (min length 1). Max length varies per model. |
negative_prompt | string, ≤ 2500–3500 | — |
duration | enum 1s..16s in 1s steps, plus 18s, 20s, 25s, 30s, 1 gen, Auto | Required. Model-specific subset. 1 gen means one generation unit for models priced per generation rather than per second. |
aspect_ratio | 1:1, 2:3, 3:2, 3:4, 4:3, 9:16, 16:9, 21:9 | Some models ignore. |
resolution | 256p..4k, or upscale hints 2x / 4x / true_1080p | Use upscale_factor for upscale models. |
upscale_factor | 1 / 2 / 4 | Only for upscale models. 1 = quality enhancement. |
audio | bool | Default true. Audio-capable models. |
image_url | URL or data: URL | Image-to-video reference frame. |
end_image_url | URL or data URL | End frame / transition reference. |
audio_url | URL or data URL | Background music input. WAV/MP3, ≤ 30 s, ≤ 15 MB. |
video_url | URL or data URL | Video-to-video / upscale input. MP4/MOV/WebM. |
reference_image_urls[] | array of URLs, ≤ 9 | Character / style consistency images. |
reference_video_urls[] | array of URLs, ≤ 3 | Reference-to-video models (e.g. Seedance 2.0 R2V). Inherits subject motion, camera movement, and style. Per clip 2–15 s, .mp4 or .mov, ≤ 50 MB; aggregate ≤ 15 s. |
reference_audio_urls[] | array of URLs, ≤ 3 | Donor audio for vocal timbre, narration, or sound effects. Per clip 2–15 s, .wav or .mp3; aggregate ≤ 15 s. Must be paired with at least one reference image or reference video — audio-only Reference workflows are rejected at validation. |
consents | object | Provider-specific consent attestations. Seedance requires consent only when the submitted media contains faces. |
elements[] | array, ≤ 4 | Advanced models (e.g. Kling O3 R2V): each has frontal_image_url, up to 3 reference_image_urls, video_url. Reference in prompt as @Element1, @Element2. |
scene_image_urls[] | array of URLs, ≤ 4 | Advanced scene refs; reference in prompt as @Image1, @Image2. |
Common recipes
Text → video with audio
{
Read the complete source ↗