RLTasksby ulam.ai
Models / Anthropic

Anthropic / MODEL EVIDENCE

Claude 4.5 Opus

27 benchmarks · 29 observations · 2 sources

Scores at a glance

27 benchmarks · large dot: Claude 4.5 Opus; small dots: closest models where reported

IFEval ↗90.9
90.994.8

Claude 4.5 Opus · IFEval · Qwen report ↗

MCPMark ↗42.3
29.557.5

Claude 4.5 Opus · MCP-Mark · original release · Qwen report ↗

MMMU ↗80.7
80.687.2

Claude 4.5 Opus · MMMU · Qwen report ↗

CharXiv ↗68.5
66.182.1

Claude 4.5 Opus · CharXiv(RQ) · Qwen report ↗

ALE-Bench ↗1025.38 Performance points
137.782951.32

claude-4.5-opus · Average contest performance ↗

Closest models by score

Similar positions on shared benchmark leaderboards. Smaller gaps are closer.

45.4 percentile-point gap

25 of 27 shared benchmark comparisons

49.0 percentile-point gap

25 of 27 shared benchmark comparisons

How similarity is measured

Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.

Filter comparisons to this model →

Reports featuring this model

Qwen3.5 · earlier-2026 baseline · Qwen →

Sources

Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.

  1. Ale Bench — methodology and published results ↗

    Epoch AI · 2026

  2. Qwen/Qwen3.5-397B-A17B · Hugging Face ↗

    Qwen Team

Names used in source tables

Claude 4.5 Opus · claude-4.5-opus