Integration
Route LiteLLM through RemKey
LiteLLM is the multi-provider LLM proxy and SDK. Here's the one change that puts it behind a governed gateway.
LiteLLM already abstracts providers, but it routes, it doesn't govern. Point its api_base at RemKey and you keep LiteLLM's ergonomics while gaining the security layer underneath: guardrails, classifier tiering, BYOK, and a tamper-evident audit chain.
Why route it through RemKey
- LiteLLM decides which provider; RemKey decides whether the request is safe and which tier it deserves, the governance LiteLLM leaves to you.
- One fail-closed PII + injection screen under every provider LiteLLM might pick, so policy doesn't fork per backend.
- BYOK with instant revoke, bring your Anthropic/OpenAI/OpenRouter keys; RemKey prefers first-party and uses OpenRouter as commodity overflow.
- A single signed audit chain across everything LiteLLM sent, regardless of which provider served it.
Swap these lines
Python SDK, set api_base per call, or globally
import litellm
response = litellm.completion(
model="openai/gpt-4o",
messages=[{"role": "user", "content": "hi"}],
api_base="https://remkey.ai/v1",
api_key="rk_live_...your RemKey key...",
)
# LiteLLM proxy (config.yaml), same idea per model:
model_list:
- model_name: gpt-4o
litellm_params:
model: openai/gpt-4o
api_base: https://remkey.ai/v1
api_key: rk_live_...your RemKey key...
Because RemKey speaks the OpenAI wire format, LiteLLM treats it as a normal openai/* provider, no custom provider class required.
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.