Product

AI usage tracker

Track usage metrics across multiple LLMs and coding assistants (Cursor, Copilot, OpenAI, and Claude) with custom AI metrics your VP Eng can share with finance.

Missing a tool or feature?

Or view the live demo · run a free audit

What the tracker measures

One dashboard for IDE assistants and API keys. No proxy, no browser plugin: official vendor APIs only.

Cursor usage

Tab completions, agent runs, tokens, and spend from the Cursor Analytics API.

GitHub Copilot metrics

Acceptance rate, active users, and seat utilization from Copilot Metrics.

Multiple LLM APIs

OpenAI and Anthropic org usage in one view: model mix, tokens, and cost.

Daily auto-sync

Sync now or every day at 06:00 UTC, with email reports when a pull completes.

Custom AI metrics without another spreadsheet

Most teams start with spend and DAU. Then they need cost per developer, idle Copilot seats, model mix, and which repos burn agent tokens. Those are the custom AI metrics ForgeMeter builds from synced usage, not a one-off export.

When you need the number for a board deck, generate an executive audit or keep a live workspace on daily sync.

  • Cost per developer across Cursor, Copilot, and APIs
  • Seat utilization vs. people who actually ship with AI
  • Model mix: Opus vs Sonnet vs GPT vs cheaper defaults
  • Spike and budget alerts before month-end invoices

AI usage tracker FAQ

Engineering tool usage, not production traces.

What is an AI usage tracker for engineering teams?+

An AI usage tracker records how developers use coding assistants and LLM APIs (seats, tokens, models, and cost) so leaders can see adoption and spend in one place. ForgeMeter tracks Cursor, GitHub Copilot, OpenAI, and Anthropic via official APIs.

How do I track usage metrics across multiple LLMs?+

Connect each provider's usage or admin API, normalize tokens and dollars into one schema, then review per-developer and per-model breakdowns. ForgeMeter does that on a daily sync so you are not exporting CSVs from four consoles.

Can I define custom AI metrics?+

Yes. Start with spend, tokens, DAU, and Copilot acceptance rate, then slice by team, repo, and model. Custom AI metrics in ForgeMeter are built from the same synced usage, not a separate spreadsheet.

Is this the same as Langfuse or Helicone?+

No. Those tools observe traces in production applications. An engineering AI usage tracker measures IDE assistants and team API keys, the tools developers use to write software.

Start tracking this week

Free workspace with usage sync. Connect one integration and see live metrics in minutes.

Missing a tool or feature?