Claude Sonnet 4vsGPT-5.2

Claude Sonnet 4
GPT-5.2
Benchmarks
Nonsense detection
BullshitBench v2
30%
38%Best
Coding
SWE-Bench Verified
72.7%
80%Best
Abstract reasoning
ARC-AGI-2
52.9%
Science
GPQA Diamond
75.4%
92.4%Best
Community preference
Arena Elo (Text)
1476
Community preference (code)
Arena Elo (Code)
1419
Overview
CompanyAnthropicOpenAI
Release dateMay 22 2025Dec 11 2025
AccessProprietaryProprietary

Which is better: Claude Sonnet 4 or GPT-5.2?

GPT-5.2 leads Claude Sonnet 4 on 3 of the 3 benchmarks they both report (BullshitBench v2, SWE-Bench Verified, GPQA Diamond). Claude Sonnet 4 shipped 203 days before GPT-5.2, 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, GPT-5.2 leads at 38% vs Claude Sonnet 4 at 30%. On SWE-Bench Verified, GPT-5.2 leads at 80% vs Claude Sonnet 4 at 72.7%. On GPQA Diamond, GPT-5.2 leads at 92.4% vs Claude Sonnet 4 at 75.4%.

Frequently asked questions

Claude Sonnet 4 was released by Anthropic on May 22 2025.

GPT-5.2 was released by OpenAI on Dec 11 2025.

GPT-5.2 leads on SWE-Bench Verified — Claude Sonnet 4 72.7% vs GPT-5.2 80%.

GPT-5.2 leads on GPQA Diamond — Claude Sonnet 4 75.4% vs GPT-5.2 92.4%.

Other comparisons