Anthropic·released 2026-09-221 source
Claude Opus 5.5 benchmark scores: 20 benchmarks tracked, leading 9. 0% of its scores are independently reproduced — the rest are self-reported or unverified.
Leads on 9 of 20 benchmarks·best result 100 on OTIS Mock AIME 2024-2025 (unverified)·0 of 20 independently reproduced
Head-to-headClaude Opus 5.5 vs GPT-6 AstraLeads on: ARC-AGI, DTBench, FrontierCode, Furniture Assembly, LMCA, MirrorCode, OTIS Mock AIME 2024-2025, ProofBench, Remote Labor Index
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| OTIS Mock AIME 2024-2025 | 100accuracy (%) | unverifiedT2 | 2026-09-22 |
| ProofBench | 100accuracy (%) | unverifiedT1 | 2026-09-22https://www.vals.ai/benchmarks/proof_bench |
| ARC-AGI | 98.5accuracy (%) | unverifiedT1 | 2026-09-22https://arcprize.org/leaderboard |
| DTBench | 98.22accuracy (%) | unverifiedT1 | 2026-09-22https://conceptualreasoning.ai/dtbench |
| FrontierMath-Tier-4-v2-Private | 95accuracy (%) | unverifiedT2 | 2026-09-22 |
| ARC-AGI-2 | 93.33accuracy (%) | unverifiedT2 | 2026-09-22 |
| FrontierMath-Tiers-1-3-v2-Private | 91.23accuracy (%) | unverifiedT2 | 2026-09-22 |
| GPQA diamond | 87.46accuracy (%) | unverifiedT2 | 2026-09-22 |
| LMCA | 80.27accuracy (%) | unverifiedT1 | 2026-09-22https://conceptualreasoning.ai/lmca |
| MirrorCode | 77.4accuracy (%) | unverifiedT2 | 2026-09-22 |
| Furniture Assembly | 76.19accuracy (%) | unverifiedT2 | 2026-09-22 |
| APEX-Agents | 73.5accuracy (%) | unverifiedT2 | 2026-09-22 |
| SimpleQA Verified | 72.2accuracy (%) | unverifiedT2 | 2026-09-22 |
| EBR-bench | 71.43accuracy (%) | unverifiedT2 | 2026-09-22 |
| Mystery Game Puzzles | 68.05accuracy (%) | unverifiedT2 | 2026-09-22 |
| Chess Puzzles | 65.28accuracy (%) | unverifiedT2 | 2026-09-22 |
| FrontierSWE | 62.33accuracy (%) | unverifiedT1 | 2026-09-22https://www.frontierswe.com/ |
| FrontierCode | 54.64accuracy (%) | unverifiedT1 | 2026-09-22https://cognition.com/frontiercode |
| Remote Labor Index | 21.25accuracy (%) | unverifiedT2 | 2026-09-22 |
| SpireClimb | 3.33accuracy (%) | unverifiedT2 | 2026-09-22 |
Claims drawn from cited facts, not live model generation.
This model was developed by Anthropic and released in 2026.
2 cited facts
This model has scores on 20 tracked benchmarks and currently leads 9 of them. The record’s evidence status is unverified.
3 cited facts
Its mean gap to the best recorded score across its tracked benchmarks is 4.08 points.
1 cited fact
1 facts cross-checked across data sources: 1 single-source.
Source facts, citations, and refresh stamp for this record.
Sources: epoch_benchmarks
Claude Opus 5.5 is an AI model developed by Anthropic, released 2026-09-22. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.
Claude Opus 5.5 has recorded scores on 20 benchmarks, each shown with its evidence status.
Claude Opus 5.5 has recorded scores on 20 benchmarks — GPQA diamond, OTIS Mock AIME 2024-2025, DTBench, LMCA, SimpleQA Verified, Chess Puzzles, and 14 more. The full table above shows each score with its evidence status.
0 of 20 recorded scores (0%) are independently reproduced rather than self-reported by the lab.
Claude Opus 5.5 has 20 tracked claims: 20 unverified.