Integration
Route Open WebUI through RemKey
Open WebUI is a self-hosted chat UI that talks to any OpenAI-compatible backend. Here's the one change that puts it behind a governed gateway.
Open WebUI is a self-hosted chat frontend that speaks the OpenAI API. It already lets you add any OpenAI-compatible connection, so it drops onto RemKey with zero plugins. You add one connection and every model your team picks from the dropdown is routed, screened, and logged.
Why route it through RemKey
- Per-connection config, not per-request headers. RemKey carries routing and policy on the key itself, so you configure it once in Open WebUI's connection settings, you never edit request bodies or inject custom headers the way a header-driven gateway forces you to.
- Classifier routing means the cheap prompts your team fires all day stop paying premium prices, an LLM classifier picks the tier per request (96% accuracy vs 33% for keyword rules).
- Every message is PII-screened (email/SSN/phone/card/API-key) and injection-screened with Llama Guard, fail-closed, before it reaches a provider.
- The whole conversation history is hash-chained and Ed25519-signed, a tamper-evident audit export your security team can verify offline.
Swap these lines
Add a RemKey connection (Settings → Connections → OpenAI API)
# Open WebUI: Admin Panel → Settings → Connections → + Add Connection # Or set these environment variables on the Open WebUI container: OPENAI_API_BASE_URL=https://api.openai.com/v1 OPENAI_API_BASE_URL=https://remkey.ai/v1 OPENAI_API_KEY=rk_live_...your RemKey key... # Model names map to RemKey tiers. Leave the model list as-is: # RemKey's classifier re-routes each request to the right tier for you.
That is the entire change. Users keep the same model dropdown; RemKey decides the real tier per message and records every one.
Drop-in by design. RemKey accepts both Authorization: Bearer and x-api-key, and serves the native OpenAI and Anthropic wire formats. The base-URL swap is the only code change.