Integrations
Connect frameworks, SDKs, coding agents and automation tools to TokenReply. Each page lists the exact setting, a minimal example and what to verify.
- LangChainUse LangChain ChatOpenAI with a TokenReply key, a catalog model ID, and a custom OpenAI-compatible base URL.
- LiteLLMSend LiteLLM Python calls to TokenReply with an OpenAI-compatible api_base while keeping model IDs explicit.
- Cline and Roo CodeConfigure Cline or Roo Code with TokenReply as an OpenAI-compatible provider, then verify tool calling on the chosen model.
- DifyConnect Dify through an OpenAI-compatible provider configuration while treating Dify plugin and model support as version-specific.
- n8nCall TokenReply from n8n with an HTTP Request node when the installed OpenAI credential does not expose a custom API base URL.
- Anthropic SDKUse the Anthropic SDK against TokenReply native Messages routing without accidentally duplicating the /v1 path.
Apps with their own guide
- JanitorAIUse the maintained JanitorAI proxy setup guide.
- SillyTavernUse the maintained SillyTavern setup and model-selection guide.
- Claude CodeUse the canonical Claude Code documentation already maintained in Docs.
- CodexUse the canonical Codex configuration documentation.
- OpenAI SDKUse the canonical Python, JavaScript, curl, and streaming examples.