When was Claude Sonnet 4.6 released?

Claude Sonnet 4.6 was released by Anthropic on Feb 17 2026.

When was GPT-5.2 released?

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

Which is better on GPQA Diamond, Claude Sonnet 4.6 or GPT-5.2?

GPT-5.2 scores higher on GPQA Diamond — Claude Sonnet 4.6 89.9% vs GPT-5.2 92.4%.

Which is better at coding, Claude Sonnet 4.6 or GPT-5.2?

On SWE-Bench Verified (real-world software-engineering tasks), GPT-5.2 leads — Claude Sonnet 4.6 scores 79.6% and GPT-5.2 scores 80%.

Home Analytics Contact

Latest Release

Muse Spark

Apr 8 2026

Home/Compare/Claude Sonnet 4.6 vs GPT-5.2

Claude Sonnet 4.6 vs GPT-5.2

0 vs 2 benchmarks won

	Anthropic Claude Sonnet 4.6	OpenAI GPT-5.2
Overview
Company	Anthropic	OpenAI
Release date	Feb 17 2026	Dec 11 2025
Model type	—	—
Open source	No	No
Specifications
Parameters	—	—
Context window	—	—
Benchmarks
Science reasoning GPQA Diamond	89.9%	92.4%
Software engineering SWE-Bench Verified	79.6%	80%
Multimodal understanding MMMU	—	—
Timeline
Release gap	GPT-5.2 shipped 68 days before Claude Sonnet 4.6

Which is better: Claude Sonnet 4.6 or GPT-5.2?

GPT-5.2 leads Claude Sonnet 4.6 on 2 of the tracked benchmarks (GPQA Diamond, SWE-Bench Verified, MMMU). GPT-5.2 shipped 68 days before Claude Sonnet 4.6, 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 GPQA Diamond, GPT-5.2 scores 92.4%, 2.5 points above Claude Sonnet 4.6 at 89.9%. On SWE-Bench Verified, GPT-5.2 scores 80%, 0.4 points above Claude Sonnet 4.6 at 79.6%.

Frequently asked questions

When was Claude Sonnet 4.6 released?: Claude Sonnet 4.6 was released by Anthropic on Feb 17 2026.
When was GPT-5.2 released?: GPT-5.2 was released by OpenAI on Dec 11 2025.
Which is better on GPQA Diamond, Claude Sonnet 4.6 or GPT-5.2?: GPT-5.2 scores higher on GPQA Diamond — Claude Sonnet 4.6 89.9% vs GPT-5.2 92.4%.
Which is better at coding, Claude Sonnet 4.6 or GPT-5.2?: On SWE-Bench Verified (real-world software-engineering tasks), GPT-5.2 leads — Claude Sonnet 4.6 scores 79.6% and GPT-5.2 scores 80%.