Claude Opus 4.6
Anthropic's strongest model for coding, agents, and frontier reasoning, with a 1M context window.
Pricing
Input$5 / 1M checked Jul 19, 2026
Cached inputSee Anthropic pricing page
Output$25 / 1M checked Jul 19, 2026
Prompt caching and batch discounts can reduce cost materially.
Cost bandPremium
Free tierChat access requires paid Claude plans; API is paid.
Limits
Context window1M checked Jul 19, 2026
Max output128k
Knowledge cutoffMay 2025
DeploymentClaude Platform, Bedrock, Vertex AI, Foundry
OpenAI-compatible APINo
ModalitiesText, Image input
Where it fits
Best forHard coding tasks, long-horizon agents, and premium research or enterprise workflows.
Business fitBuy it when output quality and agent persistence matter more than speed or price.
Capabilities
Extended thinkingAdaptive thinkingTool useVisionPriority tier
Benchmark scores
Composite93 / 100
Coding96 / 100 (SWE-bench Verified)
Reasoning95 / 100 (GPQA Diamond)
Long context91 / 100 (RULER 1M)
Vision83 / 100 (MMMU)
Instruction following95 / 100 (IFEval)
Editorial aggregate of published benchmark results as of Feb 1, 2026. These are not independent measurements run by this site.
Change history
- Feb 5, 2026New modelClaude Opus 4.6 released 1M context, $5/$25 per 1M. Extended + adaptive thinking; sets a new ceiling for coding agents.
Data
Last checkedJul 19, 2026
Cross-checked againstopenrouter.ai/anthropic/claude-opus-4.6
Verify with provider documentation before committing spend. Report an error and it gets fixed with a logged correction.