// side by side · same rules, same day
Claude Code vs Codex
Claude Code 5/5 · Codex 5/5
| Claude CodeAnthropic AGENT | CodexOpenAI AGENT | |
|---|---|---|
| CAPABILITYyes / no / uncertain | 5/5 AGENT | 5/5 AGENT |
| PROOFhow well you can check us | TRACE VERIFIED | OPERATOR INSPECTABLE |
| TRACE ACCESSwho can read the record | OPERATOR-OWNED | OPERATOR-OWNED |
| LEASHtool calls between human turns | 15 median · max 92 | |
| CHECKEDversion and date | docs as of 2026-08 (CLI 2.1.x) | Codex CLI + cloud, docs as of 2026-08 |
| CONFIDENCEours, in this dossier | 9/10 | 8/10 |
| COMMUNITYwhat visitors ruled | A0 · W0 · C0 | A0 · W0 · C0 |
| IN COURTopen challenges | ||
| DISPUTEDwhere the axes disagree |
02
Question by question
rows marked "they differ" are where the comparison earns its keep| Claude CodeAnthropic AGENT | CodexOpenAI AGENT | |
|---|---|---|
| LOOPSruns until done | ✓ yes TRACE VERIFIED Docs' own example: 'write tests for the auth module, run them, and fix any failures' from one prompt. | ✓ yes OPERATOR INSPECTABLE OpenAI docs describe one agent loop that reasons, calls tools and runs tests, iterating until they pass. |
| CHOOSESpicks its own tools | ✓ yes TRACE VERIFIED Picks tools at runtime from Bash/Read/Edit plus any MCP servers; --allowedTools only bounds the set. | ✓ yes OPERATOR INSPECTABLE The model issues shell commands and edits at runtime; sandbox_mode only bounds what those calls may touch. |
| ACTSchanges the world | ✓ yes TRACE VERIFIED Docs: it 'edits files, runs commands'; it also stages git changes, commits and opens pull requests. | ✓ yes OPERATOR INSPECTABLE Default sandbox is workspace-write: it writes files and executes shell commands in the workspace. |
| RECOVERSfixes its own errors | ✓ yes TRACE VERIFIED That same documented task ends in fixing the failures it just ran; docs also show root-cause bug fixing. | ✓ yes OPERATOR INSPECTABLE It reads command and test output in-loop and retries; docs' guidance is that it iterates until tests pass. |
| UNSUPERVISEDnobody watching | ✓ yes TRACE VERIFIED Four of the 42 scored sessions ran to completion with zero operator interventions, and non-interactive claude -p with --allowedTools runs in CI with no approval prompts. | ✓ yes OPERATOR INSPECTABLE approval_policy 'never' (and --yolo) removes prompts; Codex cloud runs whole tasks in an isolated container. |
Every answer here is the same one on the dossier and in the dataset — this page holds no numbers of its own. Think one is wrong? Bring a trace.