Skip to main content
Use the verified integrations below with RunInfra’s supported OpenAI-shaped endpoints. Configure https://api.runinfra.ai/v1, your RunInfra API key, and a deployed model ID.

LangChain

ChatOpenAI(openai_api_base=..., openai_api_key=...) with a deployed chat model ID.

LlamaIndex

Configure OpenAI(api_base=..., api_key=...) for documented LLM and embedding examples.

Vercel AI SDK

createOpenAICompatible({ baseURL, apiKey }) in your backend route.

​
The one-line rule

base_url  ->  https://api.runinfra.ai/v1
api_key   ->  YOUR_RUNINFRA_API_KEY

​
What’s not in this section

  • The OpenAI SDK itself is documented under OpenAI compatibility. This section is for higher-level libraries on top of it.
  • Observability and MLOps (Langfuse, Helicone, Arize, Datadog) pick up RunInfra traffic transparently when configured against an OpenAI-compatible endpoint. Point their base URL at RunInfra.
  • IDE extensions (editor assistants that call RunInfra directly) are not supported out of the box today. If your editor supports a custom OpenAI-compatible endpoint, configure the RunInfra base URL and key for supported endpoint fields, but we have not verified any specific editor and don’t document one until we do.

​
Missing an integration?

If you need a first-party adapter or cookbook for something that isn’t here, open an issue at RightNow-AI/runinfra-docs.