If you already call OpenAI from your language of choice, you can point the same client at Wren R1 without rewriting your call sites. Set the base URL to https://webwright.ai/v1, set the model to wren-r1, and the request you already write works.

Pick your stack:

  1. curl and Python: point an OpenAI-compatible client at Wren R1. Get a key, set the base URL, and send your first chat or transcription request. Read the guide
  2. Node.js and TypeScript: use the official openai package against Wren R1 for chat, streaming, and transcription. Read the guide
  3. Go: use the official openai-go SDK and set the base URL as a configuration value. Read the guide
  4. .NET and C#: use the official OpenAI .NET SDK and set the client endpoint. Read the guide
  5. Java: use the official OpenAI Java SDK and set the base URL and model. Read the guide
  6. Rust: use the async-openai crate and set the base URL and model. Read the guide
  7. LangChain: point ChatOpenAI at Wren R1 as a configuration change, not a rewrite. Read the guide
  8. Vercel AI SDK: point the AI SDK provider at Wren R1 with a base URL and model id change. Read the guide
  9. Ruby: use the official openai gem and set the base URL and model. Read the guide
  10. PHP and Laravel: use the community openai-php/client package and set the base URI and model. Read the guide
  11. Evaluating a provider: point your LLM evaluation framework at Wren R1. Read the guide

Every guide uses the same endpoint, the same model id wren-r1, and the same key. Chat and transcription share both, so the endpoint stays a value your deployment can change without touching call sites.

The WebWright Team