Claude Sonnet 4.6

Anthropic · Claude 4 · frontier · released Feb 17, 2026 · claude-sonnet-4-6

Claude 4 family:Claude Haiku 4.5·Claude Opus 4.6·Claude Sonnet 4.6·Claude Opus 4.8

Anthropic's best speed-to-intelligence balance for agents, coding, and professional work at scale.

Pricing

Input$3 / 1M checked Jul 19, 2026
Cached inputSee Anthropic pricing page
Output$15 / 1M checked Jul 19, 2026

Prompt caching and batch discounts apply.

Cost bandPremium-balanced
Free tierAvailable in Claude chat; API is paid.

Limits

Context window1M checked Jul 19, 2026
Max output64k
Knowledge cutoffAug 2025
DeploymentClaude Platform, Bedrock, Vertex AI, Foundry
OpenAI-compatible APINo
ModalitiesText, Image input

Where it fits

Best forTeams that need near-flagship quality with faster latency and better price efficiency than Opus.
Business fitThe likely default Anthropic buy for product and platform teams.

Capabilities

Extended thinkingAdaptive thinkingTool useVisionPriority tier

Benchmark scores

Composite86 / 100
Coding90 / 100 (SWE-bench Verified)
Reasoning84 / 100 (GPQA Diamond)
Long context88 / 100 (RULER 1M)
Vision81 / 100 (MMMU)
Instruction following88 / 100 (IFEval)

Editorial aggregate of published benchmark results as of Feb 1, 2026. These are not independent measurements run by this site.

Change history

Data

Last checkedJul 19, 2026
Cross-checked againstopenrouter.ai/anthropic/claude-sonnet-4.6

Verify with provider documentation before committing spend. Report an error and it gets fixed with a logged correction.

Open in the comparison table JSON API