openai/codex vs anthropics/claude-code
Flagged as competitors because both are in ai-dev-tools and were mentioned together in 7 threads, and appeared on the same trending list 1 round. Same-owner projects are never paired here, because they are usually pieces of one tool rather than rivals.
Stars over time
Both repositories on one timeline: same axis, same 30-day period. A line that starts later means we have less history for that repository, not that it did not exist.
Growth
Stars gained in each period. The bracketed figure is the growth percentage, which favours small repositories: 10 new stars moves a 200-star project 5%, but a 20,000-star one only 0.05%.
Today(0.1% (codex) vs 0.1% (claude-code))
Last 7 days(1.0% (codex) vs 1.1% (claude-code))
Last 30 days(3.8% (codex) vs 3.9% (claude-code))
Last year(178.7% (codex) vs 295.7% (claude-code))
Scale
What each repository is right now, in the same fields as the header on its own page, so the two are read the same way.
Stars
Forks
Open issues
Discussions
Commits
Commits landing on the default branch, per week, from GitHub's own statistics.
Merged pull requests
Pull requests merged per week: reviewed work, not raw activity.
Activity and open work
Weekly totals cover only the 30 days we have measured and state how many weeks that is. A shorter window is not a slower repository.
Commits measured(data covers 52 weeks and none)
Merged pull requests measured(data covers 2 weeks and none)
Median wait from opening to merging(400 PRs and none sampled)
Open issues (latest daily sample)
Fit
The overlap between the two: the part that says whether they answer the same question at all.
“Not measured” means the pipeline has no figure for that field yet; it is not zero. Compare a different pair · openai/codex · anthropics/claude-code.