Vision-language

Claude 3 Sonnet

By Anthropic, in the Claude Sonnet family.

Released
4 Mar 2024
Status
Retired
Retired
21 Jul 2025
Context
200K
Max output
4.1K
Input per 1M tokens
$3.00
Output per 1M tokens
$15.00
Open weights
No
License
proprietary
Parameters
—
Knowledge cutoff
Aug 2023
Reasoning
No
Modalities
text, image → text

Category scores

Out of 100, from independent and third-party results only, by method category-2026-09-29.1.

Category scores
CategoryScoreRankCoverage
Human preference21.7Not ranked2 of 5

Benchmark results

Every result we hold for this model, with who ran it and where it was published.

Knowledge & reasoning

Knowledge & reasoning results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
DTBench (Decision Theory Benchmark)Accuracy (capability questions)53.6%Third party——https://conceptualreasoning.ai/dtbench via Epoch AI
MMLUAccuracy75.9%Third party——Stanford CRFM Leaderboard via Epoch AI
WinoGrandeAccuracy75.1%Self-reported——The Claude 3 Model Family: Opus, Sonnet, Haiku via Epoch AI

Scientific reasoning

Scientific reasoning results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
GPQA DiamondAccuracy40.6%Independent—27 Jan 2025Epoch AI

Mathematics

Mathematics results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
MATH Level 5Accuracy18.2%Independent—27 Jan 2025Epoch AI
OTIS Mock AIME 2024-2025Accuracy2.5%Independent—25 Feb 2025Epoch AI

Coding & software engineering

Coding & software engineering results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
WeirdML (v2)Average accuracy10.2%Third party——https://htihle.github.io/weirdml.html via Epoch AI

Human preference

Human preference results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Text Arena (arena.ai)Arena score (Elo scale)1281Independent—25 Sep 2026Arena
Text Arena (arena.ai)Votes109284Independent—25 Sep 2026Arena
Vision Arena (arena.ai)Arena score (Elo scale)1016Independent—28 Sep 2026Arena
Vision Arena (arena.ai)Votes12314Independent—28 Sep 2026Arena

Other Claude Sonnet versions

Sources

Claude 3 Sonnet: benchmarks, price and specs · supermarket.dev