Claude Opus 4.5

AnthropicReleased 24 Nov 2025ClosedFlagshipDate verified

Compare Announcement

153142–164Index score#50 of 229 models

7/ 8Core tests taken5 of 6 areas · plus 21 anchor tests

−8Versus the frontier at releasebest then: Gemini 3 ProReplaced by Claude Opus 4.6

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

200KContext windowtokens it reads at once

Where it sitson the frontier

GPT-5 reached this score 4 months earlier.

Test resultsby area · tick = best by any model

Reasoning & knowledgeevidence

GPQA Diamond

86%

Humanity's Last Exam

25%

Mathsevidence

FrontierMath T1-3

34%

Codingevidence

SWE-bench Verified

77%

Terminal-Bench

63%

Agentsevidence

APEX-Agents

no result

Novel problemsevidence

ARC-AGI-2

38%

Long tasksevidence

METR time horizon

4.9 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#50of 219 · 153

Epoch Capabilities Indexoverall capability from many benchmarks#48of 239 · 150.1 LiveBench · Overalloverall, with fresh questions to limit contamination#43of 58 · 72.6

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.