blockrun_video
Generate short AI videos from text or image prompts using multiple models, paying only for successful renders.
Instructions
Generate short AI videos via BlockRun x402 (async, client-polled).
Turns a text prompt (and optional seed image) into a short MP4 clip. The tool submits the job, then polls until the video is ready (typical total wall-time 60-180s; 5 min hard cap). Payment is settled only when upstream returns a finished video — if the job fails or we give up, you are not charged.
Models. Every rate below is what you are CHARGED (margin and transaction fee included), at the 720p baseline Seedance renders by default with synced audio:
azure/sora-2 (~$0.105/sec, 720p + synced audio, text-to-video) — OpenAI Sora 2 via Azure AI Foundry. duration_seconds must be 4, 8, or 12 (4s default -> ~$0.42/clip). No image_url / RealFace.
xai/grok-imagine-video ($0.05/sec at 480p default, $0.07/sec at 720p; 8s default -> $0.401/clip, 1-15s) — stylized, fast. 480p/720p only.
bytedance/seedance-1.5-pro (~$0.071/sec, 4-12s, 5s default -> ~$0.35/clip) — cheapest Seedance, token-priced upstream
bytedance/seedance-2.0-mini (~$0.080/sec, 4-15s, 5s default) — 2.0-generation quality at roughly half the 2.0-fast rate; 720p ceiling; supports RealFace and first/last-frame
bytedance/seedance-2.0-fast (~$0.165/sec, 4-15s, ~60-80s gen) — sweet-spot price/quality; supports BytePlus RealFace assets
bytedance/seedance-2.0 (~$0.227/sec, 4-15s, up to 4K) — highest quality, and the ONLY model that renders true 4K; supports RealFace, first/last-frame and reference media
bytedance/seedance-2.5 (~$0.315/sec, 4-30s, 5s default) — long-form: double 2.0's length ceiling, multilingual. NOT a strict upgrade — it caps at 720p and does NOT support RealFace or first/last-frame. Use 2.0 for 1080p/4K or real-person video.
Image-to-video is NOT cheaper than text-to-video on Seedance — same per-second rate. Higher resolutions ARE more expensive (token-priced: 1080p ~2.25x, 4K ~9x the 720p rate); the 402 quote is authoritative and is what gets charged.
RealFace: to generate video of a SPECIFIC real person, first enroll them with blockrun_realface (returns a ta_xxxx asset id), then pass real_face_asset_id here with seedance-2.0, seedance-2.0-fast, or seedance-2.0-mini. Mutually exclusive with image_url.
Returns a permanent blockrun-hosted MP4 URL (the gateway mirrors the asset to GCS so URLs don't expire).
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| model | No | Video model to use | xai/grok-imagine-video |
| prompt | Yes | Text description of the video to generate. E.g. 'a red apple slowly spinning on a wooden table', 'a hummingbird hovering near a red flower, ultra slow motion' | |
| agent_id | No | Agent identifier for budget tracking and enforcement. | |
| image_url | No | Optional seed image URL for image-to-video generation | |
| resolution | No | Output resolution. Seedance defaults to 720p and is token-priced (~2.25x at 1080p, ~9x at 4K); per-model sets from token360's published schema: seedance-2.0 480p/720p/1080p/4K · 1.5-pro 480p/720p/1080p · 2.0-fast, 2.0-mini and 2.5 480p/720p only. grok-imagine-video honours 480p (default, $0.05/sec) and 720p ($0.07/sec) and rejects anything higher. Ignored by Sora only (dropped from the request). | |
| aspect_ratio | No | Output aspect ratio. Seedance honors the full set; Sora uses it only to pick portrait vs landscape (9:16 / 3:4 -> portrait); Grok ignores it (the gateway never forwards it to xAI). Defaults to the model's own default. (9:21 removed 2026-08-07 — no Seedance model offers it; use 9:16 for vertical.) | |
| generate_audio | No | Seedance only: whether to generate a synced audio track. Defaults ON for text-to-video and OFF for image/RealFace-conditioned. The auto-generated audio is occasionally rejected by upstream moderation ('output audio may contain sensitive information') even for benign prompts — pass false to skip audio and avoid that failure. Ignored by xAI/Sora. | |
| last_frame_url | No | Seedance 1.5-pro / 2.0 / 2.0-fast / 2.0-mini only (NOT 2.5): first-and-last-frame interpolation. A second image URL that seeds the FINAL frame so the model tweens from image_url (first frame) → last_frame_url (last frame). Requires image_url; mutually exclusive with real_face_asset_id. | |
| duration_seconds | No | Duration to bill for. Defaults to the model's own default (8s xAI, 5s Seedance, 4s Sora). Per-model range: seedance-1.5-pro 4-12s · seedance-2.0 / 2.0-fast / 2.0-mini 4-15s · seedance-2.5 4-30s · sora-2 exactly 4, 8 or 12 · grok-imagine-video 1-15s. | |
| real_face_asset_id | No | BytePlus RealFace asset id (from blockrun_realface enroll/list) to generate video of a specific real person. Seedance 2.0 / 2.0-fast / 2.0-mini only (NOT 2.5). Mutually exclusive with image_url. |