Video Generation
HappyHorse Series
Alibaba Cloud Bailian HappyHorse Video — single model, unified entry, intelligent routing across T2V/I2V/R2V/EDIT
POST
HappyHorse Series
Video generation model on Alibaba Cloud Bailian.
Single model, unified entry — just one
After submission, poll status with
happyhorse-1.0. The backend automatically routes to T2V / I2V / R2V / EDIT based on the fields you pass, no model id switching needed.
Billed by resolution × duration, any integer between 3-15 seconds.
Pricing
Mode Routing
Examples
GET /v1/tasks/{task_id}. See Task System for details.
Available Models
Common Parameters
string
required
Fixed value
happyhorse-1.0string
default:"generate"
Operation type. Usually no need to pass explicitly — the backend auto-routes based on media fields. Allowed values:
generate— text-to-video (T2V)image2video— image-to-video (I2V), used withfirst_frame_imagereference— reference-to-video (R2V), used withimage_urlsedit— video editing, used withvideo_urlreference_audio— used withaudio_urlsfor audio-driven generation
string
Video description, up to 5000 characters. Required for T2V; optional as guidance for other modes
integer
default:"5"
Video duration in seconds, any integer in 3-15. In EDIT mode,
0 means follow the source video’s durationstring
default:"720p"
Output resolution. Allowed values:
720p1080p
string
default:"16:9"
Frame aspect ratio. Only effective in T2V; in I2V / R2V / EDIT it follows the source media. Allowed values:
16:9— landscape widescreen9:16— portrait tall4:3— landscape3:4— portrait1:1— square
string[]
Reference image URL array, 1-9 images:
- Triggers R2V mode (multi-image style / character consistency)
- In EDIT mode, layered as style references (≤5 images)
string
First-frame image URL, triggers I2V mode
string[]
Source video URL array (single element), triggers EDIT mode:
- base64 not currently supported
- Must be a publicly downloadable URL
string[]
Reference audio URL array (
reference_audio mode)Model-Specific Parameters
string
default:"auto"
Audio setting, only effective in EDIT mode. Allowed values:
auto— automatically generate audio matching the edited videoorigin— keep the source video’s original audio track
Resource Limits
Related Docs
- Task System — task state machine / polling cadence / async push
- Request & Response Format — common error codes / Headers / rate limits
- Authentication — API Key issuance and usage
HappyHorse Series