Guides

n8n

n8n's built-in OpenAI credential is built for OpenAI's own models. Point its base URL at Tempr Gateway and that same credential — the one your workflows already use — can call Claude, Gemini, DeepSeek, or anything else on your allowlist, with no separate credential type per provider.

Why this matters

n8n's "OpenAI" credential is normally exactly that — OpenAI only. Once it points at Tempr, the model field on each node becomes the only thing that decides which provider actually runs, so one workflow can mix models from different providers across its nodes without juggling multiple credential types.

  1. Create a virtual key

    In the Portal, create a Gateway virtual key. Keys are prefixed tvk_ and shown once at creation.

  2. Add a provider key

    Gateway is BYOK: add a key for at least one provider from the supported list.

  3. Add an OpenAI credential in n8n

    Go to Settings → Credentials → Add Credential → OpenAI. Set:

    • API Key: your virtual key
    • Base URL: https://api.temprhq.io/v1
    • Organization ID: leave blank

  4. Use it in a workflow

    Add an AI Agent node with a Chat Model edge using this credential, and set the model field to any id from your allowlist — anthropic/claude-sonnet-4-5, google/gemini-3-pro, whatever the workflow needs, not just an OpenAI model.

Note

Toggle off the Responses API in the node — Tempr Gateway implements the Chat Completions API (/v1/chat/completions), not OpenAI's newer Responses API.

What's next

Previous← OpenCode NextGitHub Copilot App →