The contenders
The coding agents under test.
Each agent works in a terminal on its own repository with the same spec and time budget. Results belong to a specific run and configuration, not to the product forever. Tools change fast.
- Claude CodeAnthropic
Terminal coding agent
Anthropic's agentic coding tool. It works in the terminal, reads the codebase, runs commands and makes changes on its own.
1 recorded run
- OpenAI CodexOpenAI
Terminal coding agent
OpenAI's coding agent. It plans, edits and runs code from the command line, inside a sandbox with configurable approval policies.
1 recorded run
- Cursor CLICursor
Terminal coding agent
The command-line agent from the makers of the Cursor editor. It runs the Cursor agent headless in any terminal.
1 recorded run
No aggregate leaderboard is shown. Cross-experiment statistics will appear only once there is enough independently verified data to support them, with tool versions and configurations recorded per run.