Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models741Deals30d268Policy30d67, 15 pendingSources1,014

Claude Sonnet 4.6

AnthropicReleased Feb 17, 2026Current

Score basis

Claude Sonnet 4.6 is strong in long context; capable in coding; and behind the leaders in reasoning, agentic tasks, factuality, multimodal tasks, math, and instruction following. Too few results yet to rate safety or multilingual tasks.

Capability profile

Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.

Strong1 capability
  1. Long Context−5.7%2 of 3
Capable1 capability
  1. Coding−22.5%8 of 10
Limited6 capabilities
  1. Reasoning−25.2%5 of 6
  2. Agentic−27.8%7 of 7
  3. Factuality−28.0%4 of 4
  4. Multimodal−33.8%2 of 6
  5. Math−39.9%1 of 5
  6. Instruction Following−42.1%2 of 3
Not rated2 capabilities

Too few results yet to rate Safety or Multilingual.

Price

$3.00input$15.00outputper million tokens

From Artificial Analysis · 3 providers tracked · All prices

Costs more than 88% of 331 priced models · 3:1 input-to-output blend, log scale

Evidence

152results on88benchmarks

  • 24 independently verified
  • 52 aggregator
  • 45 vendor-reported
  • 31 cross-referenced

From 20 sources · latest Oct 8, 2026 · How verification works

API features

Tool callingStructured outputsReasoning

As listed by OpenRouter

Research

102 papers reference Claude Sonnet 4.6

Claude Sonnet 4.6 benchmark results

152 results on 88 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.

Long Context

StrongFull long context ranking

5.7% behind the leader2 of 3 ranked benchmarks measured

Show 2 more long context resultsHide 2 long context results

Coding

CapableFull coding ranking

22.5% behind the leader8 of 10 ranked benchmarks measured

Show 8 more coding resultsHide 8 coding results

Reasoning

LimitedFull reasoning ranking

25.2% behind the leader5 of 6 ranked benchmarks measured

Show 13 more reasoning resultsHide 13 reasoning results

Agentic

LimitedFull agentic ranking

27.8% behind the leader7 of 7 ranked benchmarks measured

Show 11 more agentic resultsHide 11 agentic results

Factuality

LimitedFull factuality ranking

28.0% behind the leader4 of 4 ranked benchmarks measured

Show 6 more factuality resultsHide 6 factuality results

Multimodal

LimitedFull multimodal ranking

33.8% behind the leader2 of 6 ranked benchmarks measured

Show 7 more multimodal resultsHide 7 multimodal results

Math

LimitedFull math ranking

39.9% behind the leader1 of 5 ranked benchmarks measured

Show 1 more math resultHide 1 math result

Instruction Following

LimitedFull instruction following ranking

42.1% behind the leader2 of 3 ranked benchmarks measured

Show 2 more instruction following resultsHide 2 instruction following results

More results

Benchmarks outside the capability baskets. They are not ranked against a leader.

Show 66 more resultsHide 66 results

Claude Sonnet 4.6: common questions

Who makes Claude Sonnet 4.6?

Claude Sonnet 4.6 is made by Anthropic.

When was Claude Sonnet 4.6 released?

Claude Sonnet 4.6 was released on Feb 17, 2026, according to Artificial Analysis.

What is Claude Sonnet 4.6 good at?

Claude Sonnet 4.6 is strong in long context; capable in coding; and behind the leaders in reasoning, agentic tasks, factuality, multimodal tasks, math, and instruction following. Too few results yet to rate safety or multilingual tasks.

How much does Claude Sonnet 4.6 cost?

Claude Sonnet 4.6 costs $3.00 per million input tokens and $15.00 per million output tokens, according to Artificial Analysis. We track its price at 3 providers. At a mix of three input tokens to one output token, it costs more than 88% of the 331 priced models we track.

How many benchmarks has Claude Sonnet 4.6 been tested on?

We track 152 results for Claude Sonnet 4.6 on 88 benchmarks from 20 sources, 24 of them independently verified. The latest was recorded on Oct 8, 2026.

Which API features does Claude Sonnet 4.6 support?

OpenRouter lists tool calling, structured outputs, and reasoning for Claude Sonnet 4.6.

About this record

Where Claude Sonnet 4.6's numbers come from, and every name it appears under.

Tracked since
Apr 27, 2026
Newest source mention
Sep 10, 2026

Where the results come from

Verification: 152 scores · 24 independently verified · 52 aggregator-attributed · 31 vendor cross-reference · 45 vendor-reported. How these tiers are assigned

From 20 sources on 13 sites. Artificial Analysis supplies 52 of them; the 24 independently verified results come from 7 sites. Bars are coloured by trust tier.

  • artificialanalysis.ai52
  • www-cdn.anthropic.com29
  • deepmind.google22
  • api.llm-stats.com16
  • huggingface.co8
  • livebench.ai7
  • arcprize.org4
  • epoch.ai4
  • raw.githubusercontent.com4
  • datasets-server.huggingface.co2
  • lmarena.ai2
  • labs.scale.com1
  • mistral.ai1

Also known as

How our sources name Claude Sonnet 4.6 at each reasoning setting.

SettingShort formLong formAPI id
lowclaude sonnet 4.6 (non-reasoning, low)claude sonnet 4.6 (non-reasoning, low effort)claude-sonnet-4-6-non-reasoning-low-effort
highclaude sonnet 4.6 (high) claude sonnet 4.6 (non-reasoning, high)Claude Sonnet 4.6 (Non-reasoning, High Effort)—
maxSonnet 4.6 Thinking (Max) claude sonnet 4.6 (max)Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)—
Also listed assonnet-4-6sonnet 4.6claude-sonnet-4-6claude sonnet 4.6 (16k thinking)claude sonnet 4.6 (32k thinking)claude-sonnet-4.6 (non-reasoning)sonnet~4.6claude-sonnet-4.6 (Reasoning)claude 4.6 sonnet