v0.3.0

OpenAPI 3.1.0

Submit a video generation task

Felo API Support

MIT

服务器

https://openapi.felo.ai

客户端库

Shell Curl

Submit a video generation task

Creates an async video generation task and returns a task_id immediately (HTTP 202). Poll GET /api/v1/videos/generations/{task_id} until it succeeds. The request body can be provided either flattened ({model, prompt, ...}) or nested ({model, input: {prompt, ...}}); the input object takes precedence when present. callback fields (callBackUrl / callback_url / callbackUrl) are ignored.

Three mutually exclusive input modes:

text-to-video: prompt only;
image-to-video (first/last frame): first_frame_url (+ optional last_frame_url);
omni-reference: reference_image_urls / reference_video_urls / reference_audio_urls. Mixing first/last-frame with any multimodal reference input returns 400 invalid_request_error.

请求头

X-Request-Id
类型: string

Optional caller request ID echoed back for diagnostics.

请求体·VideoGenerationRequest

必填

application/json

Video generation input. Either flattened ({model, prompt, ...}) or nested ({model, input: {...}}). input takes precedence. Exactly one input mode must be used; mixing first/last-frame with multimodal reference inputs is rejected.

model
类型: string
必填

Felo-facing model name, e.g. seedance-2.0-mini.
aspect_ratio
类型: string 枚举
值
- 16:9
- 9:16
- 1:1
- 4:3
- 3:4
- 21:9
- adaptive
duration
类型: integer
最小值:
4
最大值:
15

Integer numbers.
first_frame_url
类型: string 格式: uri
generate_audio
类型: boolean
input
类型: object

Optional nested input object; when present, its fields take precedence over sibling fields.
- aspect_ratio
  类型: string 枚举
  
  Output aspect ratio; adaptive follows the input image ratio.
  
  值
  16:9
  9:16
  1:1
  4:3
  3:4
  21:9
  adaptive
- duration
  类型: integer
  最小值:
  4
  最大值:
  15
  
  Output duration in seconds.
- first_frame_url
  类型: string 格式: uri
  
  First frame image URL (image-to-video mode).
- generate_audio
  类型: boolean
  
  Whether to generate a synchronized audio track.
- last_frame_url
  类型: string 格式: uri
  
  Last frame image URL; must be used together with first_frame_url.
- prompt
  类型: string
  最大长度:
  2000
  
  Prompt. Required for text-to-video and omni-reference modes.
- reference_audio_urls
  类型: array string[] …3格式: uri
  
  Omni-reference audio clips.
- reference_image_urls
  类型: array string[] …9格式: uri
  
  Omni-reference images. Mutually exclusive with first/last-frame.
- reference_video_urls
  类型: array string[] …3格式: uri
  
  Omni-reference videos. Mutually exclusive with first/last-frame. Supplying video changes the billing unit price and adds input duration to the token count.
- resolution
  类型: string 枚举
  
  Output resolution. Higher resolution consumes more tokens and costs more.
  
  值
  480p
  720p
  1080p
  4k
last_frame_url
类型: string 格式: uri
prompt
类型: string
最大长度:
2000
reference_audio_urls
类型: array string[] …3格式: uri
reference_image_urls
类型: array string[] …9格式: uri
reference_video_urls
类型: array string[] …3格式: uri
resolution
类型: string 枚举
值
- 480p
- 720p
- 1080p
- 4k

响应

202
类型: object · VideoGenerationAccepted

Task accepted
kind
枚举
常量:
video
必填
值
video
outputs
类型: array string[]
必填

Empty until the task succeeds.

status
枚举
常量:
queued
必填
值
queued
task_id
类型: string
必填

Task id to use for polling.
application/json
400
类型: object · GenerationError

Invalid request (unknown model, mixed input modes, invalid fields)

OpenAI-compatible error envelope.
error
类型: object
message
类型: string
必填
type
类型: string 枚举
必填
值
invalid_request_error
server_error
code
类型: string
param
类型: string 可为 null
application/json
401
类型: object · GenerationError

Invalid or missing Felo API key

OpenAI-compatible error envelope.
error
类型: object
message
类型: string
必填
type
类型: string 枚举
必填
值
invalid_request_error
server_error
code
类型: string
param
类型: string 可为 null
application/json
402
类型: object · GenerationError

Insufficient balance / billing requirement not met

OpenAI-compatible error envelope.
error
类型: object
message
类型: string
必填
type
类型: string 枚举
必填
值
invalid_request_error
server_error
code
类型: string
param
类型: string 可为 null
application/json
403
类型: object · GenerationError

Account not permitted (subscription required or user banned). Disabled models are also rejected before upstream calls. Existing accepted requests and tasks continue.

OpenAI-compatible error envelope.
error
类型: object
message
类型: string
必填
type
类型: string 枚举
必填
值
invalid_request_error
server_error
code
类型: string
param
类型: string 可为 null
application/json
429
类型: object · GenerationError

Request or spending limit reached

OpenAI-compatible error envelope.
error
类型: object
message
类型: string
必填
type
类型: string 枚举
必填
值
invalid_request_error
server_error
code
类型: string
param
类型: string 可为 null
application/json
502
类型: object · GenerationError

Upstream provider failed

OpenAI-compatible error envelope.
error
类型: object
message
类型: string
必填
type
类型: string 枚举
必填
值
invalid_request_error
server_error
code
类型: string
param
类型: string 可为 null
application/json
503
类型: object · GenerationError

Generation service or billing service temporarily unavailable

OpenAI-compatible error envelope.
error
类型: object
message
类型: string
必填
type
类型: string 枚举
必填
值
invalid_request_error
server_error
code
类型: string
param
类型: string 可为 null
application/json

post/api/v1/videos/generations

curl https://openapi.felo.ai/api/v1/videos/generations \
  --request POST \
  --header 'Content-Type: application/json' \
  --header 'Authorization: Bearer YOUR_SECRET_TOKEN' \
  --data '{
  "model": "seedance-2.0-mini",
  "prompt": "a cinematic trailer shot, 720p",
  "resolution": "720p",
  "duration": 5,
  "generate_audio": true
}'

显示 Schema

{
  "task_id": "string",
  "status": "queued",
  "kind": "video",
  "outputs": [
    "string"
  ]
}

Task accepted

Submit a video generation task

身份验证Required

Submit a video generation task