AI Rate Limit Tracker

Track, forecast, and schedule around AI provider rate limits — without getting 429'd.

OpenAIAnthropicGemini GroqCohereMistral DeepSeek

Get API Key

Endpoints

GET
/providers
List all supported providers + models (no auth)
GET
/limits/:provider?model=...
Published RPM/TPM/RPD/context specs (no auth)
POST
/track
Record a call — get live usage counters + warnings at 80%+ (auth required)
GET
/usage?provider=&model=
Current minute/day window counters (auth required)
GET
/forecast?provider=&rate_rpm=
Exhaustion ETA + scheduling recommendations (auth required)

Quick Start

curl https://api.lazy-mac.com/ai-rate-limit-tracker/limits/openai?model=gpt-4o # Track a call curl -X POST https://api.lazy-mac.com/ai-rate-limit-tracker/track \ -H "X-API-Key: YOUR_KEY" \ -H "Content-Type: application/json" \ -d '{"provider":"openai","model":"gpt-4o","tokens_used":1500}' # Forecast exhaustion curl "https://api.lazy-mac.com/ai-rate-limit-tracker/forecast?provider=openai&model=gpt-4o&rate_rpm=10" \ -H "X-API-Key: YOUR_KEY"

MCP Integration

{"jsonrpc":"2.0","id":1,"method":"initialize","params":{}}

POST to /ai-rate-limit-tracker/mcp — tools: get_provider_limits, list_providers, track_usage, get_forecast