A web app that tracks AI coding tool spend across Copilot, Claude Code, and Cursor, normalized per commit and per PR
GitHub Copilot's June 2026 switch from flat-rate to per-token billing caused immediate cost spikes of 10x-50x for teams, with one user burning 54% of a monthly quota on a single request and the community discussion receiving 958 downvote reactions vs 24 upvotes. Simultaneously, the Claude Code source leak exposed 512k lines of proprietary scaffolding code, triggering mass debate about AI coding tool transparency and vendor lock-in. Teams now juggle three or more AI coding tools with zero visibility into normalized spend across them, and no cross-vendor spend web app exists that ties cost to actual output like commits, PRs, and repos.
Demand Breakdown
Social Proof 3 sources
Gap Assessment
4 tools exist (GitHub Copilot Billing Dashboard, Anthropic Console Usage, Cursor Usage, CloudZero) but gaps remain: Single-vendor only, no output normalization to cost per commit or per PR, no alerting before overage, and no cross-tool view across Claude Code or Cursor.; Claude-only, no code-output attribution, no per-repo or per-developer breakdown, no connection to GitHub commits or PRs..
Features2 agent-ready prompts
Competitive LandscapeFREE
| Product | Does | Missing |
|---|---|---|
| GitHub Copilot Billing Dashboard | Shows Copilot seat usage and cost for a single GitHub organization in the GitHub billing UI. | Single-vendor only, no output normalization to cost per commit or per PR, no alerting before overage, and no cross-tool view across Claude Code or Cursor. |
| Anthropic Console Usage | Shows aggregate token counts and cost per API key for Claude models. | Claude-only, no code-output attribution, no per-repo or per-developer breakdown, no connection to GitHub commits or PRs. |
| Cursor Usage | Shows in-app usage metrics for Cursor subscriptions. | Cursor-internal only, no API export, no normalization to commits or PRs, no cross-vendor view. |
| CloudZero | Cloud infrastructure cost intelligence platform for AWS, Azure, and GCP spend optimization. | Infrastructure cost tooling only, no AI coding tool vendor billing API integrations for Copilot, Claude Code, or Cursor, no developer-output attribution. |
Notable VoicesFREE
"The 54% monthly quota burned in one request story hit 767 HN points and 553 comments. Teams have no pre-flight cost estimate for any Copilot request."
"Claude Code npm source leak thread hit 2095 HN points and 1022 comments, triggering the mass transparency debate about AI coding tool vendor lock-in."
Leads1000BUILDER
Sign in to unlock full access.