Claude Sonnet 4.6

Closed Source

Closed Weights

Parameters

Undisclosed

Context Length

Modality

Multimodal

Architecture

Undisclosed

License

Proprietary

Release Date

17 Feb 2026

Knowledge Cutoff

Aug 2025

API Pricing (per 1M)

Input: $3.00 · Output: $15.00

Evaluation Benchmarks

Rank

#51

Benchmark	Score	Rank
Coding LiveBench Coding	medium 0.79	21
Data Analysis LiveBench Data Analysis	medium 0.78	24
Graduate-Level QA GPQA	0.899	25
Reasoning LiveBench Reasoning	medium 0.85	34
General Text Text Arena	1472	40
General LiveBench Global	medium 0.73	42
Mathematics LiveBench Mathematics	medium 0.87	42
LiveBench Average LiveBench Average	medium 0.73	43
Agent Arena Agent Arena	-1.52	43
Web Development WebDev Arena	1521	44
Agentic Coding LiveBench Agentic	medium 0.43	49
Agentic Index Artificial Analysis	max 0.32	60
Coding Index Artificial Analysis	max 0.63	63
Intelligence Index Artificial Analysis	max 0.30 Standard 0.25	96 129

Rankings

Overall Rank

#51

Coding Rank

#46

About Claude Sonnet 4.6

Claude Sonnet 4.6 is a high-performance multimodal workhorse from Anthropic optimized for agentic workflows, GUI navigation, and complex codebases. It features hybrid adaptive reasoning and context compaction across a 1M token context window.

Technical Specifications

Architecture specifications are undisclosed for proprietary models.

Resources

Official Documentation Release Notes

About Claude 4

Anthropic's fourth generation Claude models with advanced reasoning, extended context windows up to 200K tokens, and configurable thinking effort levels. Features improved safety alignment, nuanced understanding, and sophisticated task completion. Includes Opus (most capable), Sonnet (balanced), and Haiku (fast) variants, with thinking modes that enable transparent chain-of-thought reasoning for complex problems.