Skip to content
aiqbench

Anthropic · model passport

Claude Opus 4

claude-opus-4
  • Cloud only
  • Proprietary licence
  • Released 2025-05-22(curated)
13/27axes covered
Overall · LMArena
1,426#107 of 396
Best percentile
83.8Instruction following
Cohort standing
53.3median 50.3

What the data says

Benchmarked on 13 of 27 tracked axes. Every figure below is the model's own published result on that board.

Uneven: 56.2 percentile points between its best and worst axis — 83.8 pct on Instruction following versus 27.6 pct on HHEM Hallucination (Vectara).

  • Instruction following83.8 pct1,444Verify →
  • Coding81.6 pct1,498Verify →
  • Hard prompts76.3 pct1,457Verify →
  • Overall73.2 pct1,426Verify →
  • SimpleBench67.8 pct58.8%Verify →
  • GPQA Diamond56.6 pct76.3%Verify →
  • Vision Arena49.3 pct1,208Verify →
  • AIME (math)45.5 pct64.4%Verify →
  • ARC-AGI-242.7 pct8.6%Verify →
  • Humanity's Last Exam39.1 pct10.7%Verify →
  • FrontierMath31.9 pct4.5%Verify →
  • SWE-bench Verified28.1 pct70.7%Verify →
  • HHEM Hallucination (Vectara)27.6 pct88.0%Verify →
  • WebDev ArenaNot benchmarked
  • Factuality ArenaNot benchmarked
  • Search ArenaNot benchmarked
  • Document ArenaNot benchmarked
  • Agent ArenaNot benchmarked
  • SimpleQA VerifiedNot benchmarked
  • LiveBenchNot benchmarked
  • BFCL v4 (function calling)Not benchmarked
  • SWE-bench Verified (live board)Not benchmarked
  • Terminal-Bench 4.0Not benchmarked
  • MMLU-Pro (Artificial Analysis)Not benchmarked
  • LiveCodeBench (Artificial Analysis)Not benchmarked
  • IFBench (Artificial Analysis)Not benchmarked
  • Terminal-BenchNot benchmarked

Percentiles are the model's standing within each board's own population. Axes marked "not benchmarked" have no published result for this model — that is an absence of data, not a low score.

  • claude-opus-4-20250514-thinking-16k stands at 53.3 (raw mean 54.1) — measured on 13 of 27 axes.

    standingtemplateVerify →

  • claude-opus-4-20250514-thinking-16k stands at 53.3 versus a cohort median of 50.3 across 146 models released within 180 days — above the peer median.

    cohorttemplateVerify →

Every claim is a computed fact from the current board data, phrased by a low-temperature model call and rejected by a validation gate if any number is not in the source row. How the leaderboards work →