Claude Sonnet 4.6

Ideal balance of intelligence and speed

Claude Sonnet 4.6 strikes the sweet spot between capability and cost. It delivers near-Opus performance on most tasks at a fraction of the price, making it the recommended model for most production applications.

Try Claude Sonnet 4.6

Specifications

Context window200k tokens

Input price$3.00 / 1M tokens

Output price$15.00 / 1M tokens

Credits per query3 cr

Released2025-01

LicenseProprietary

Features

VisionTool useJSON modeFunction callingCode

Benchmark Scores

What do these mean?

Sourced from official model cards and academic papers. Higher is better.

GPQA Diamond

PhD-level science & reasoning

61%

AIME 2024

Mathematical olympiad problems

52%

Arena Elo

Chatbot Arena human preference ranking

1312

HumanEval

Python coding — pass@1 rate

93%

MATH

Competition-level mathematics

80%

MMLU

General knowledge across 57 academic domains

89%

MT-Bench

Multi-turn instruction following (out of 10)

—

SWE-Bench

Real-world GitHub issue resolution rate

70.3%

More from Anthropic

Claude Opus 45 cr

Anthropic's most intelligent model

Claude 3.7 Sonnet3 cr

Extended thinking for deep reasoning

Claude Haiku 4.62 cr

Anthropic's fastest and most affordable model