One API for every model.
LightRouter routes your requests to AI models through one endpoint and one bill, with transparent per-token pricing.
Quickstart
curl https://api.lightrouter.ai/v1/chat/completions \
-H "Authorization: Bearer $LIGHTROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "lightrouter/mock",
"messages": [{"role": "user", "content": "Hello"}]
}'Access is by invitation while we are in private beta.
How we operate
Authorized supply only
Every model is served through a channel its provider permits for resale. No shared keys, no pooled accounts.
Prices you can read
Per-token prices are listed for every model and returned by the API. You are charged for the tokens the provider reports, nothing else.
Drop-in compatible
The API follows the OpenAI chat-completions format, streaming included. Point your existing SDK at a new base URL.