OpenAI-API-compatible
Verified by DifyModel providers compatible with OpenAI's API standard, such as LM Studio or llama.cpp Service.
Overview
This plugin provides access to models that are OpenAI-compatible, including LLMs, reranking, text embedding, speech-to-text (STT), and text-to-speech(TTS) models. Developers can easily add models by providing configuration parameters such as the model name and API key.
Configure
Configure the OpenAI-API-compatible model by providing its core details (Type, Name, API Key, URL) and adjusting further options like completion, context, and token limits, as well as streaming and vision settings. Save when done.
TokenLab example
TokenLab can be configured through this plugin with the OpenAI-compatible API surface:
Use for LLM/chat models. For non-LLM model types where this plugin appends the API version internally, such as text embedding, rerank, STT, and TTS, use instead to avoid a duplicated path.
Other TokenLab chat models can be added the same way, for example , , , , , , , , , , and .
TokenLab also provides native endpoint families such as Responses, Anthropic Messages, and Gemini-compatible APIs. This OpenAI-API-compatible plugin uses TokenLab's OpenAI-compatible endpoint.
Some outbound calls build their address at runtime, so the list may be incomplete.
Private network address — this plugin accesses an internal network host.
Private network address — this plugin accesses an internal network host.
Last checked 10/07/2026 11:56 AM · version 0.0.68