Video Editing
Creates an asynchronous video editing task (multi-video stitching) using an OpenAI-compatible request. The editing endpoint is routed to the video editing model pool (models that support both video generation and video editing); Doubao-Seedance-2.5 is in that pool.
Unlike the generation endpoint, at least one media input is required: input_reference / first_frame / last_frame / input_video. input_video is only available on this endpoint and references public video URLs for stitching, as a single string or an array; it can be freely combined with prompt and image inputs. All other fields (model / prompt / size / duration / parameters) are identical to the generation endpoint.
The response and the polling / download endpoints are shared with generation tasks.
Authentication
Use Authorization: Bearer $JOY_TOKEN_API_KEY.
Request
Model name, e.g. Doubao-Seedance-2.5. Omit to use auto and let the gateway route within the video generation model pool.
Text description. Required for text-to-video; may be omitted when any image input (input_reference / first_frame / last_frame) is present.
Resolution/aspect. Accepts 480p / 720p / 1080p or WIDTHxHEIGHT (e.g. 1280x720, converted to the aspect ratio and resolution). Aspect ratios support 16:9 and 9:16; use parameters.aspect_ratio for other ratios.
Video length in seconds. Continuous 2~30 range (Seedance 2.5), default 8. Out-of-range values are clamped (<2 becomes 2, >30 becomes 30). Stitch longer videos through POST /openai/v1/videos/edits.
Reference image URL(s). A single string is one image; an array is multiple images for multi-subject reference (each sent with the reference_image role). Up to 30 images per request (input_reference + first_frame + last_frame combined), otherwise 400.
First-frame image URL (first_frame role).
Last-frame image URL (last_frame role).
Optional end-user identifier for abuse tracking.
Native upstream parameter passthrough; snake_case field names are preferred (camelCase accepted). Supports aspect_ratio, resolution, duration_seconds, generate_audio, seed, negative_prompt, compression_quality, and person_generation. fps, sample_count, and 4K resolutions are not supported (rejected or ignored). Top-level duration / size explicitly provided take precedence.
Public URL(s) of the video(s) to stitch, only available on the editing endpoint. A single string is one video; an array is multiple videos. Freely combinable with prompt and image inputs. Must be a directly accessible public URL.
Response
The actually selected model name (the real model after auto routing).
Unix timestamp (seconds) of task creation.
