claude code vs cursor vs codex — measured, not surveyed
| # | Agent | Devs | Tokens | Share | Top model | Est cost |
|---|---|---|---|---|---|---|
| 01 | Codex CLIcodex | 161 | 864.56B | 46% | gpt-5.6-sol | $561,849 |
| 02 | Claude Codeclaude | 169 | 722.08B | 38% | claude-opus-4-8 | $736,954 |
| 03 | Cursorcursor | 92 | 175.94B | 9.3% | default | $132,497 |
| 04 | OpenCodeopencode | 90 | 79.32B | 4.2% | unreported | $57,104 |
| 05 | Factory Droiddroid | 10 | 10.58B | 0.6% | unreported | $4,056 |
| 06 | Hermeshermes | 39 | 10.06B | 0.5% | unreported | $5,540 |
| 07 | Openai Apiopenai-api | 3 | 6.63B | 0.4% | gpt-realtime-1.5 | $1,972 |
| 08 | Openclawopenclaw | 17 | 5.57B | 0.3% | unreported | $2,266 |
| 09 | Pipi | 27 | 5.51B | 0.3% | unreported | $1,771 |
| 10 | Kimi CLIkimi | 5 | 3.15B | 0.2% | k3 | $991 |
| 11 | Roo Coderoo | 3 | 2.38B | 0.1% | roo | $1,157 |
| 12 | Gemini CLIgemini | 43 | 1.96B | 0.1% | unreported | $601 |
| 13 | Anthropic Apianthropic-api | 1 | 980.4M | <0.1% | claude-sonnet-4-6 | $1,358 |
| 14 | Ampamp | 9 | 875.6M | <0.1% | unreported | $992 |
| 15 | GitHub Copilotcopilot | 1 | 412.2M | <0.1% | gpt-4.1 | $1,403 |
| 16 | Kilo Codekilo | 7 | 262.7M | <0.1% | unreported | $4.79 |
| 17 | Clinecline | 4 | 259.4M | <0.1% | cline | $1.16 |
| 18 | Antigravityantigravity | 1 | 57.0M | <0.1% | gemini 3.5 flash (low) | $228 |
| 19 | Qwen Codeqwen | 5 | 45.2M | <0.1% | coder-model | $0 |
| 20 | Grokgrok | 1 | 5.3M | <0.1% | grok-3 | $7.2 |
| 21 | Goosegoose | 3 | 733.1K | <0.1% | unreported | $2.91 |
| 22 | Continuecontinue | 1 | 406.5K | <0.1% | starcoder2:3b | $0 |
npx whoburnedmore CLI reads the local usage logs coding agents already keep and submits daily aggregate token totals per tool and model — never prompts, code or file names. Panel = 203 listed developers who opted in publicly, including subscription usage (Claude Code, Cursor, Codex) that API-gateway rankings never see. Costs estimated from public per-token pricing. what we collect · leaderboard · state of AI coding