All times Eastern (ET)Provider pages checked daily at
ProSign inRSS

Claude Sonnet 4.6

API ID claude-sonnet-4-6

claude-sonnet-4-6 is an Anthropic model, released February 2026. It has a 1,000,000-token context window and up to 128,000 output tokens. It takes image input and supports reasoning. Standard price: $3.00 input and $15.00 output per 1M tokens.

Status
Active
Released
February 2026
Shutdown
not before Feb 17, 2027
Replacement
—

claude-sonnet-4-6 is active. Anthropic says it won't be retired before February 17, 2027.

Benchmarks

Scores as each source published them
or

Comparing with Claude Sonnet 4.5, the previous model in this line. A score is compared only with the same benchmark from the same source, at the same setting where both have it. Full comparison: prices, limits, features

Arena

ArenaClaude Sonnet 4.6Claude Sonnet 4.5ChangeRankDateSource
Text
By category
  • Math 1463 rank 62 vs 1428 (+35)
  • Creative writing 1449 rank 40 vs 1454 (−5, within range)
  • Instruction following 1474 rank 29 vs 1462 (+12)
  • Hard prompts 1503 rank 26 vs 1483 (+20)
  • Long prompts 1492 rank 26 vs 1482 (+10, within range)
  • Multi-turn 1481 rank 37 vs 1478 (+3, within range)
  • Expert 1506 rank 34 vs 1483 (+23)
1472 ±41455 ±3+1745 of 415
70,940 votes		Oct 13, 2026source
Text: Coding1528 ±61512 ±5+1624 of 410
19,727 votes		Oct 13, 2026source
WebDev1522 ±51385 ±7+13749 of 142
22,884 votes		Oct 13, 2026source
Vision1275 ±6not rated34 of 165
27,422 votes		Oct 9, 2026source
Search1202 ±51176 ±5+2617 of 42
173,986 votes		Oct 13, 2026source
Document1478 ±61449 ±6+2913 of 44
57,813 votes		Sep 13, 2026source

Rated as claude-sonnet-4-6; Claude Sonnet 4.5 as claude-sonnet-4-5-search. Ratings come from people voting between two anonymous answers; Text, Vision, Search and Document are style-controlled, as on arena.ai. Data: Arena (arena.ai), CC BY 4.0.

Epoch AI

BenchmarkClaude Sonnet 4.6Claude Sonnet 4.5ChangeSettingRunSource
GPQA Diamond
Other settings
  • 83.3% vs not run high effort
  • 83.3% vs not run medium effort
  • 78.8% vs not run max effort
87.4%81.7%+5.7 pts32kFeb 20, 2026source
OTIS Mock AIME 2024-2025
Other settings
  • 82.2% vs not run medium effort
  • 75.6% vs not run high effort
  • 71.1% vs not run max effort
85.8%77.8%+8.0 pts32kFeb 20, 2026source
SimpleQA Verified
Other settings
  • 32.8% vs not run max effort
35.5%30.7%+4.8 ptshigh effort (vs 59k)Aug 10, 2026source
SWE-bench Verified75.2%71.3%+3.9 ptsFeb 21, 2026source

Run independently by Epoch AI. Data: Epoch AI, Capabilities & benchmarking (epoch.ai/benchmarks), CC BY.

On cloud platforms

All platform dates
PlatformPlatform IDStateShutdownTime leftReplacementSource
Amazon Bedrockanthropic.claude-sonnet-4-6Activenot before Feb 17, 2027—source
Microsoft Foundryversion 1claude-sonnet-4-6ActiveFeb 10, 2027118 days—source

Read daily from each platform's own pages. The shutdown date and countdown on this page use Anthropic's own date. Time left counts to 00:00 UTC on the shutdown date, which is the evening before in ET.

Features

Compare models
FeatureValueSinceSource
Prompt cachingA prompt prefix you send again and again (system prompt, documents, tools) is cached, so later requests cost less and start faster.Supportedin place when tracking began (October 6, 2026)source
Batch APISend many requests as one job that runs in the background, usually finished within 24 hours, at a lower price than live requests.Supportedin place when tracking began (October 6, 2026)source
Image inputImages such as screenshots, photos and charts can go in the prompt for the model to read.Supportedin place when tracking began (October 6, 2026)source
Reasoning / thinkingThe model can work through a problem step by step before it answers: usually better answers to hard questions, at the cost of more tokens and time.Supported (thinking: adaptive, extended deprecated; default effort: high)in place when tracking began (October 6, 2026)source
Context windowThe most the model can work with in one request, in tokens: your prompt, conversation history and files. A token is about three-quarters of an English word.1,000,000 tokensin place when tracking began (October 6, 2026)source
Max output tokensThe longest reply the model can write in one response, in tokens. A token is about three-quarters of an English word.128,000 tokensin place when tracking began (October 6, 2026)source

Not published by Anthropic: Structured outputs, Function calling, PDF input, Fine-tuning, Built-in web search, temperature parameter, top_p parameter, logprobs.

Price history

USD per 1M tokens, standard tier
$0.00$5.00$10.00$15.00$20.00Oct 2026In $3.00Out $15.00Oct 1, 2026 to Oct 15, 2026: Input $3.00, Output $15.00
$0.00$5.00$10.00$15.00$20.00Oct 2026In $3.00Out $15.00Oct 1, 2026 to Oct 15, 2026: Input $3.00, Output $15.00

We started recording this model's prices on Oct 1, 2026. The history grows with each daily check.

Show as a table
FromTypePrice
Oct 1, 2026Input$3.00
Oct 1, 2026Output$15.00

Prices

USD per 1M tokens, every type and band, by tier
TypePrice $/1MIn effect
Standard
input$3.00
cached input$0.30
cache write (5 min)$3.75
cache write (1 hour)$6.00
output$15.00
Batch
input$1.50
output$7.50

Changelog

Changes we detected on Anthropic's pages

No changes recorded yet.

Email alerts

Get an email if claude-sonnet-4-6 changes

We'll email you the same morning we spot a deprecation notice, a new or moved shutdown date, a price change, or the model leaving Anthropic's model list.

No password: one email to confirm, and you can unsubscribe any time. Privacy

Last checked · Sources: Anthropic models, Anthropic deprecations, Anthropic pricing, Anthropic model pages