Claude Sonnet 4.6
FrontierAnthropic•Released on 2026-02-17
Anthropic's most capable Sonnet yet. 1M context window (beta), 30-50% faster than Sonnet 4.5, approaching Opus-level intelligence at 1/3 the cost. Default model on claude.ai. Excels at coding, computer use, agent planning, and long-context reasoning.
88
Overall Score
Voice of the community
sample 1,520“Early testers preferred Sonnet 4.6 over Sonnet 4.5 roughly 70% of the time. Even more surprising, users chose it over Opus 4.5 59% of the time.”
“Approaching Opus-level intelligence at a fraction of the cost. Human-level computer use capability.”
“Sonnet 4.5 was like a 9/10 in creative writing. 4.6 is closer to a 3, even with a proper prompt. No way the quality drop should have been that hard.”
Core Specs
1000K
Context Window
128K
Max Output
ReasoningOpen Sourcetextimage
Scenario Scores
Pros & Cons
Sentiment65% +18% ·17% −
Pros
- +1M context window (beta)
- +Near-Opus intelligence at Sonnet price
- +30-50% faster than Sonnet 4.5
- +Human-level computer use capability
- +Excellent for agentic workflows
Cons
- −1M context still in beta
- −Some users report inconsistent reasoning
- −Still not Opus for most complex tasks
Pricing
Input (per 1M tokens)$3.00
Output (per 1M tokens)$15.00
Subscription$20/month
Free trial available
Updated on 2026-03-07
Get Started with Claude Sonnet
💡 Sonnet 4.6 is free to use on claude.ai.
Benchmarks
sweBench80%
osWorld72%
mmlu91.2%
humanEval95.5%
Tools Supporting This Model
Reliability
SLA99.9%
Incidents (30d)7
Last Incident2026-06-01
6/1 Anthropic 单模型多发日: 5 起单模型 resolved 事件 (Opus 4.7 06:48-07:22 34min + 12:15-13:33 78min / Sonnet 4.5 09:08-09:28 20min / Sonnet 4.6 14:40-15:41 61min + 18:28-19:06 38min)，error 签名全天在 4.5/4.6/4.7 间轮转，均 minor/单模型/无 Telegram
View Status Page →