Content, sources and update rules

The goal of this set of documentation is to help users verify configurations, not to replace technical descriptions with unverifiable blurbs.

Content scope

Covers OpenAI Compatible API, SSE streaming output, production launch, cost control, development tool access, common status codes, model selection and API Key security. The tutorial gives priority to providing the minimum reproducible request and troubleshooting sequence.

fact check

For client fields, please refer to the public documentation of the corresponding product first; for HTTP status codes, refer to the standard RFC; for AIFast addresses, models and authentication methods, refer to the API documentation and console.

Dynamic information

Models, prices, contexts, packages, and client APIs are subject to change. The document does not write the dynamic list as a permanent commitment, and is ultimately subject to the real-time console and API return.

How we write a tutorial

First give the scope of application and the prerequisites of the agreement.
Verify authentication, address, and model using minimal requests.
Distinguish between basic chat, streaming output, tool invocation, and multimodal capabilities.
Specifies the status code, response body, and request ID that should be saved on failure.
Attached are verification dates and public sources.

Documentation does not make promises

There is no promise that all models will be permanently available or that prices will never change.
Don't write "OpenAI Compatible" as "fully compatible with all OpenAI features."
Agents, tool calls, and long tasks are all available without inferring the success of a single chat.
It is not recommended to save API Keys in public code, front-end pages, or screenshots.

An error was found in the content

Please pass AIFast customer service page Provide specific pages, error fields, client versions, and reproducible information. When it comes to dynamic models and prices, please also provide the console screenshot time.