Getting Started
SDK setup, init() options, and how to send your first event. Up and running in under 60 seconds.
Read on GitHubResources
Everything you need to integrate the Model Meter SDK and understand how it works.
SDK setup, init() options, and how to send your first event. Up and running in under 60 seconds.
Read on GitHubFull event schema, init() options, OTel span processor setup, flush() for serverless, and per-operation labeling.
Read on GitHubIngestion endpoint, analytics endpoints, authentication, request/response format, and rate limits.
Read on GitHubHow the SDK, API, worker, cost engine, and dashboard fit together. Data flow from request to dashboard.
Read on GitHubModel Meter vs. Helicone, LangSmith, Langfuse, and OpenMeter — what each tool is for and when to choose each.
Read on GitHubTenant isolation via PostgreSQL RLS, ingestion key management, privacy by default, and production deployment controls.
Read on GitHubTutorials, cost analyses, and architectural deep-dives on AI spend and LLM observability.
ExploreTokens, caching, attribution, latency, cost per request, effective-dated pricing, and more.
ExploreThe workflow from SDK integration to live spend dashboard.
ExploreSupported providers, gateway adapters, and billing import paths.
ExploreUsage tracking, cost attribution, live dashboards, budgets, and alerts.
ExploreProvider rates and billing categories change frequently. Verify the model, service tier, token categories, and billing terms directly before estimating production spend. Model Meter's pricing catalogue uses effective-dated records so historical estimates are not rewritten when prices change.
Launch offer