API Overview

The base URL is always https://locopalai.com, and all APIs use the same platform API key. For shared conventions (authentication, response structure, pagination, rate limiting, idempotency), see Authentication and Common Conventions.

Models and generation APIs

Purpose	Method	Path	Details
Query available models	GET	`/v1/models`	Models and Text API
Text chat (including streaming)	POST	`/v1/chat/completions`	Models and Text API
Image generation	POST	`/v1/images/generations`	Image Generation API
Text-to-speech	POST	`/v1/audio/speech`	Text-to-Speech API
Create a video task	POST	`/v1/video/generations`	Video Generation API
Query a video task	GET	`/v1/video/generations/{task_id}`	Video Generation API
Create a video task (OpenAI-compatible form)	POST	`/v1/videos`	Video Generation API
Query a video task (OpenAI-compatible form)	GET	`/v1/videos/{task_id}`	Video Generation API

Asset library APIs

All asset library APIs use the API key and return a uniform success response: { "code": 0, "data": ..., "msg": "..." }.

Purpose	Method	Path	Details
Query asset groups	GET	`/api/v1/materials/groups`	Asset Groups
Create an asset group (`virtual` / `real`)	POST	`/api/v1/materials/groups`	Asset Groups
Query the asset list (supports `group_id` and pagination)	GET	`/api/v1/materials/assets`	Assets
Submit an image asset (`image_url` or `data_url`)	POST	`/api/v1/materials/assets`	Assets
Query a single asset's status	GET	`/api/v1/materials/assets/{asset_id}`	Assets
Sync asset status	POST	`/api/v1/materials/sync`	Sync Asset Status
Create a portrait QR code authorization session	POST	`/api/v1/materials/portrait/sessions`	Portrait Asset Authorization
Confirm portrait QR code completion	POST	`/api/v1/materials/portrait/resolve`	Portrait Asset Authorization

For detailed parameters, examples, and error descriptions, see Asset Library API Overview.

Typical call flows

Models and video

GET  /v1/models
  → POST /v1/video/generations
  → GET  /v1/video/generations/{task_id}

After creating a task, you must save the task ID; if the network is interrupted, keep querying the original task instead of creating a new one.

Text-to-speech

GET  /v1/models
  → POST /v1/audio/speech
  → 保存二进制音频响应

Whether the audio API is available and which voice and response_format values are supported depend on the current key's model catalog and the actual model capabilities.

Virtual assets

POST /api/v1/materials/groups
  → POST /api/v1/materials/assets
  → GET  /api/v1/materials/assets/{asset_id}   （轮询到 status 为 Active）
  → 在视频 content 中使用 asset://<asset_id>

Portrait assets

POST /api/v1/materials/groups {"type":"real"}
  → POST /api/v1/materials/portrait/sessions   （拿到二维码链接）
  → 由真人本人扫码完成采集
  → POST /api/v1/materials/portrait/resolve
  → POST /api/v1/materials/sync
  → 查询素材状态后用于支持真人素材库的模型

For an end-to-end example, see Portrait Assets and Reference Video Integration Workflow; for the capability boundaries of multiple images and reference videos, see Multiple Images and Reference Videos.

Responses and errors

API type	Success	Failure
Model / image / video	Returns the result object directly	HTTP non-`2xx`; read `error.message`
Asset library	`{ "code": 0, "data": ..., "msg": "OK" }`	HTTP status code matches `code`

Clients must check both the HTTP status and the response body. For the complete list of error codes and troubleshooting steps, see Error Codes and Troubleshooting.

On this page