What we measured (and why)GPT-5 Codex: How it solves for GPT-5's drawbacks
by David Loker
September 30, 2025
7 min read
Back to blog
Content
- What we measured (and why)
- What we looked for
- Scoreboard - Codex improves signal-to-noise
- Style and structure (why Codex reads like a patch)
- The kinds of bugs Codex is good at
- Why GPT‑5 felt noisier, and how we fixed that
- Latency: Fast matters & Codex is faster
- What a CodeRabbit user should expect
- Quantitative appendix (for the curious)
- Where Codex still needs work (and what we are doing)
- Codex GPT-5: All of the great bug catching ability, fewer downsides
Share
Cut code review time & bugs by 50%
Most installed AI app on GitHub and GitLab
Free 14-day trial
Share
Cut code review time & bugs by 50%
Most installed AI app on GitHub and GitLab
Free 14-day trial
Keep reading
Still have questions?
Contact us