Claude 3.7 SonnetvsGPT-5-Codex
Claude 3.7 Sonnet | GPT-5-Codex | |
|---|---|---|
| Benchmarks | ||
BullshitBench v2Nonsense detection — Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | 49% | 39% |
| BenchmarksPublished by one model only | ||
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | 62.3% | — |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 68% | — |
| Overview | ||
| Company | Anthropic | OpenAI |
| Release date | Feb 24 2025 | Sep 15 2025 |
| Access | Proprietary | Proprietary |
Other comparisons
Claude 3.7 SonnetvsGemini 3.7 FlashGPT-5-CodexvsGemini 3.7 FlashClaude 3.7 SonnetvsMuse GlimmerGPT-5-CodexvsMuse GlimmerClaude 3.7 SonnetvsGrok 4.6GPT-5-CodexvsGrok 4.6Claude 3.7 SonnetvsDeepSeek-V4-Pro-0813GPT-5-CodexvsDeepSeek-V4-Pro-0813Claude 3.7 SonnetvsMistral Medium 3.5GPT-5-CodexvsMistral Medium 3.5Claude 3.7 SonnetvsKimi K3GPT-5-CodexvsKimi K3
Frequently asked questions
Claude 3.7 Sonnet leads GPT-5-Codex on 1 of the 1 benchmark they both report (BullshitBench v2). Claude 3.7 Sonnet shipped 203 days before GPT-5-Codex, so benchmark comparisons should account for the intervening progress.
Published specifications for these two models are limited — see each model page for the latest details.
On BullshitBench v2, Claude 3.7 Sonnet leads at 49% vs GPT-5-Codex at 39%.
Claude 3.7 Sonnet was released by Anthropic on Feb 24 2025.
GPT-5-Codex was released by OpenAI on Sep 15 2025.