Anthropic·released 2024-02-291 source
Claude 3 Opus benchmark scores: 13 benchmarks tracked. 23% of its scores are independently reproduced — the rest are self-reported or unverified.
Leads on 0 of 13 benchmarks·best result 79.47 on MMLU (unverified)·3 of 13 independently reproduced
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| MMLU | 79.47accuracy (%) | unverified· optimizedT1 | 2024-02-29Stanford CRFM Leaderboard |
| Winogrande | 77accuracy (%) | unverified· optimizedT1 | 2024-02-29The Claude 3 Model Family: Opus, Sonnet, Haiku |
| MATH level 5 | 37.48accuracy (%) | reproduced· optimizedT1 | 2024-02-29Epoch AI |
| DTBench | 36accuracy (%) | unverifiedT1 | 2024-02-29https://conceptualreasoning.ai/dtbench |
| GPQA diamond | 29.55accuracy (%) | reproduced· optimizedT1 | 2024-02-29Epoch AI |
| METR Time Horizons | 29.48accuracy (%) | self-reportedT1 | 2024-02-29https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks/ |
| LMCA | 20accuracy (%) | unverifiedT1 | 2024-02-29https://conceptualreasoning.ai/lmca |
| WeirdML | 19.22accuracy (%) | unverifiedT1 | 2024-02-29https://htihle.github.io/weirdml.html |
| SimpleQA Verified | 12.6accuracy (%) | unverifiedT2 | 2024-02-29 |
| Cybench | 10accuracy (%) | unverified· optimizedT1 | 2024-02-29Cybench leaderboard |
| SimpleBench | 8.2accuracy (%) | unverifiedT1 | 2024-02-29SimpleBench Leaderboard |
| OTIS Mock AIME 2024-2025 | 4.63accuracy (%) | reproduced· optimizedT1 | 2024-02-29Epoch AI |
| Chess Puzzles | 0accuracy (%) | unverifiedT2 | 2024-02-29 |
Claims drawn from cited facts, not live model generation.
This model was developed by Anthropic and released in 2024.
2 cited facts
This model has scores on 13 tracked benchmarks and currently leads 0 of them. The record’s evidence status is independently reproduced.
3 cited facts
Its mean gap to the best recorded score across its tracked benchmarks is 58.7 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 3 Opus is an AI model developed by Anthropic, released 2024-02-29. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.
Claude 3 Opus has recorded scores on 13 benchmarks, each shown with its evidence status.
Claude 3 Opus has recorded scores on 13 benchmarks — Winogrande, GPQA diamond, MATH level 5, OTIS Mock AIME 2024-2025, DTBench, WeirdML, and 7 more. The full table above shows each score with its evidence status.
3 of 13 recorded scores (23%) are independently reproduced rather than self-reported by the lab.
Claude 3 Opus has 13 tracked claims: 3 independently reproduced, 1 self-reported, 9 unverified.