GitHub Copilot switches to per-token AI credit billing on June 1. If you use Copilot through VS Code, Copilot CLI, OpenCode, Pi, etc, your costs might be significantly higher than the current premium request model.
I built this for people who want to know the impact on their experience before June 1st.
It runs entirely locally, reads your local session logs, applies the published per-token pricing, and generates an HTML report comparing your usage against Pro, Pro+, Business, and Enterprise plan limits.
Some caveats:
- VS Code doesn't persist input/cache tokens, so those are estimated (output is exact)
- Copilot CLI normal chat exposes output tokens only; input is estimated (compaction events are exact)
- OpenCode and Pi expose full token breakdowns therefore the insights there are exact
GitHub Copilot switches to per-token AI credit billing on June 1. If you use Copilot through VS Code, Copilot CLI, OpenCode, Pi, etc, your costs might be significantly higher than the current premium request model.
I built this for people who want to know the impact on their experience before June 1st.
It runs entirely locally, reads your local session logs, applies the published per-token pricing, and generates an HTML report comparing your usage against Pro, Pro+, Business, and Enterprise plan limits.
Some caveats: - VS Code doesn't persist input/cache tokens, so those are estimated (output is exact) - Copilot CLI normal chat exposes output tokens only; input is estimated (compaction events are exact) - OpenCode and Pi expose full token breakdowns therefore the insights there are exact
Open to feedback and contributions!