Claude Opus 4.6

AnthropicReleased 5 Feb 2026ClosedFlagshipDate verified

Compare Announcement

167155–178Index score#26 of 229 models

8/ 8Core tests taken6 of 6 areas · plus 19 anchor tests

−3Versus the frontier at releasebest then: GPT-5.3-CodexReplaced by Claude Opus 4.7

$5.0/ $25Price per million tokensinput / output, list price

1MContext windowtokens it reads at once

Where it sitson the frontier

Among the first models to reach this score.

Test resultsby area · tick = best by any model

Reasoning & knowledgeevidence

GPQA Diamond

91%

Humanity's Last Exam

34%

Mathsevidence

FrontierMath T1-3

66%

Codingevidence

SWE-bench Verified

79%

Terminal-Bench

80%

Agentsevidence

APEX-Agents

46%

Novel problemsevidence

ARC-AGI-2

69%

Long tasksevidence

METR time horizon

12 h

Independent results from the Epoch AI benchmarking hub, best reasoning setting. The core tests are shown here; anchor tests also inform the score and can be shown above.

What others sayrank on each leaderboard

The Model Indexindependent tests, our method#26of 219 · 167

Epoch Capabilities Indexoverall capability from many benchmarks#27of 239 · 155.3 LiveBench · Overalloverall, with fresh questions to limit contamination#35of 58 · 74.5

Further left is better. Leaderboards measure different things: LMArena is people voting blind between two answers, the others are test scores. Click any row for the source.