01 · ACCESS
One gateway key
Call every connected model through one stable endpoint. Revoke and limit each consumer independently.
Edge-native LLM gateway
One API.
Every model you trust.
Connect authorized AI providers, route around failures, and control usage from one clean gateway.
OpenAI compatibleUsage meteringAutomatic failover
chat-completions.sh200 OK
curl https://murahrouter.pages.dev/api/v1/chat/completions \
-H "Authorization: Bearer rm_live_…" \
-H "Content-Type: application/json" \
-d '{
"model": "glm-5.2",
"messages": [{"role":"user","content":"Hello"}]
}' Routed successfully3.2s · 798 tokens
BUILT FOR RELIABLE AI ACCESS
Cloudflare EdgeOpenAI-compatibleEncrypted secretsTelegram Mini App
WHY MURAHROUTERLess integration work.
Less integration work.
More reliable requests.
Everything needed to operate multiple model providers without exposing upstream credentials.
02 · RELIABILITY
Priority routing and failover
Requests move to the next healthy authorized upstream when a provider returns an error.
03 · CONTROL
Know every token
See usage, latency, status, and model cost. Enforce a clear USD quota for each API key.
FLEXIBLE BY DESIGNYour model catalog,
Your model catalog,
behind one interface.
Use stable public model IDs while upstream names and providers change behind the scenes.
Public APIsmart-chatStable model ID
Smart routePriority + failoverPolicy at the edge
ProvidersAuthorized upstreamsKeys stay encrypted
QUICKSTART
From zero to first request.
- 1Add upstream
Connect an authorized OpenAI-compatible provider.
- 2Map a model
Create a public model ID and route priority.
- 3Issue a key
Set its quota, copy it once, then call the API.