Cost

See what your agents spend.

Verity meters the tokens every coding agent uses in your repositories and shows the spend per model, session, run, repository and organisation.

Example data
claude-opus$412
claude-sonnet$218
claude-haiku$47
Cost by model, last 30 days · every run also shows its own cost
  • Per repository

    Totals, trend, and spend by model, agent and session. Each run page shows what that run cost.

  • Across your org

    Usage by org groups every repository you can read by the organisation that owns it, over a window you choose.

  • From the terminal

    verity status sums up recent activity and the trend, without opening the dashboard.

Opt-in, one switch per machine

verity init asks once. Turn it on or off any time with verity telemetry install or uninstall. It uses Claude Code's own OpenTelemetry export, and only Verity projects send anything.

What is collected: token counts and estimated cost by model, session and agent, request timing, tool names, the repository name and the session id. What is not: your prompts, your code, or what tools read and returned.

Dollar figures are estimated from token counts and published model prices. Use them to compare and track spend, not to reconcile an invoice.

Read the docs

Agents write the code. Verity keeps it true.

Start on the Free plan. Working with Claude Code today.

$npm install -g @codacy/verity-cli && verity init

Works with

OpenAI CodexSoonCI/CDSoonGitHub Agentic WorkflowsSoonAI GatewaySoon