wylon

API Manual

Use unified authentication to access wylon text, image, and video generation. The manual is organized by capability and workflow so you can quickly find the correct endpoint and parameters.

Integration basics

Base URLall V1 endpoints
https://api.wylon.cn/v1
AuthorizationHTTP Bearerrequired
Create an API key in the dashboard and store it in $WYLON_API_KEY. Never place a key in frontend code or a public repository.
Protocol compatibilityvaries by endpoint
Text generation supports OpenAI Chat Completions and Anthropic Messages. Synchronous image generation is OpenAI Images-compatible. Asynchronous image and video generation uses wylon content-task endpoints.

Choose an API by capability

The catalog is organized by user goal. Image and video tasks share a path but use different type values and do not share every parameter.

Make your first read-only call

List models first to confirm the service URL, authentication, and target model ID.

cURL
curl https://api.wylon.cn/v1/models \
  -H "Authorization: Bearer $WYLON_API_KEY"

Synchronous and asynchronous calls

Synchronous image generation

The client remains connected and receives the image in the same response. Use this for interactive experiences where the user is waiting.

Asynchronous content tasks

Create a task, save its id, and query the status. Use this for background image processing and longer video generation.

HTTP 200 on task creation means the task was accepted. The query response status determines whether generation succeeded. Result URLs expire, so download or persist them promptly.

Error handling and retries

StatusCommon causeRecommended action
400Invalid field, type, or model parameterCorrect the request using the endpoint and model documentation
401Missing, invalid, or revoked API keyCheck the authorization header and key status
403Insufficient model or task accessCheck the account, key, and model permissions
429Rate, concurrency, or capacity limitReduce request frequency and increase retry intervals gradually
500, 503Internal error or temporary model unavailabilityQuery the original task first if you already have its ID
504Synchronous generation exceeded its wait timeConfirm the original request state before creating another request

Repeating a generation POST creates a separate request or task and may incur additional usage.

沪ICP备2026010432号-1 沪公网安备31010402336632号