Rankings / Anthropic
Claude Opus 4.1 20250805
released 2025-08-05
BenchAtlas Index
as of 2026-09-11
thinking
rank #107
8 families · 5 categories · high
base configuration
rank #142
8 families · 4 categories · medium
thinking
rank #173
7 families · 5 categories · medium
Benchmark evidence
68 results
external indices
| ECI | 144.1 points | independent | Epoch AI Benchmarking Hub | |
| ECI | 24K budget | 144.1 points | independent | Epoch AI Benchmarking Hub |
| ECI | 32K budget | 144.1 points | independent | Epoch AI Benchmarking Hub |
| ECI | 16K budget | 144.1 points | independent | Epoch AI Benchmarking Hub |
| ECI | 27K budget | 144.1 points | independent | Epoch AI Benchmarking Hub |
factuality
| MASK contamination=Potential contamination warning: This model was evaluated after the public release of MASK, allowing model builder access to the prompts and solutions. | thinking | 94.2 % 92.4–96.0 | independent | Scale Labs |
| MASK contamination=Potential contamination warning: This model was evaluated after the public release of MASK, allowing model builder access to the prompts and solutions. | 87.4 % 85.7–89.1 | independent | Scale Labs | |
| SimpleQA Verified | 27K budget | 34.8 % | independent | Epoch AI Benchmarking Hub 2025-12-09 |
human preference
| Coding (style control)older version arena=text · category=coding · style_control=true | 1505.5 1500.0–1510.9 | community | LMArena Leaderboard Dataset | |
| Industry Software And It Services (style control)older version arena=text · category=industry_software_and_it_services · style_control=true | 1486.6 1482.1–1491.1 | community | LMArena Leaderboard Dataset | |
| Hard Prompts English (style control)older version arena=text · category=hard_prompts_english · style_control=true | 1481.9 1476.6–1487.2 | community | LMArena Leaderboard Dataset | |
| Hard Prompts (style control)older version arena=text · category=hard_prompts · style_control=true | 1477.5 1473.5–1481.4 | community | LMArena Leaderboard Dataset | |
| Industry Medicine And Healthcare (style control)older version arena=text · category=industry_medicine_and_healthcare · style_control=true | 1472.5 1462.9–1482.1 | community | LMArena Leaderboard Dataset | |
| Longer Query (style control)older version arena=text · category=longer_query · style_control=true | 1472.2 1467.1–1477.3 | community | LMArena Leaderboard Dataset | |
| Multi Turn (style control)older version arena=text · category=multi_turn · style_control=true | 1469.0 1463.2–1474.9 | community | LMArena Leaderboard Dataset | |
| Chinese (style control)older version arena=text · category=chinese · style_control=true | 1468.4 1460.1–1476.7 | community | LMArena Leaderboard Dataset | |
| Spanish (style control)older version arena=text · category=spanish · style_control=true | 1467.1 1453.6–1480.7 | community | LMArena Leaderboard Dataset | |
| Expert (style control)older version arena=text · category=expert · style_control=true | 1465.4 1455.9–1474.9 | community | LMArena Leaderboard Dataset | |
| Industry Life And Physical And Social Science (style control)older version arena=text · category=industry_life_and_physical_and_social_science · style_control=true | 1463.7 1457.8–1469.6 | community | LMArena Leaderboard Dataset | |
| French (style control)older version arena=text · category=french · style_control=true | 1461.5 1444.5–1478.5 | community | LMArena Leaderboard Dataset | |
| Industry Legal And Government (style control)older version arena=text · category=industry_legal_and_government · style_control=true | 1458.2 1449.7–1466.8 | community | LMArena Leaderboard Dataset | |
| English (style control)older version arena=text · category=english · style_control=true | 1456.0 1451.9–1460.2 | community | LMArena Leaderboard Dataset | |
| Instruction Following (style control)older version arena=text · category=instruction_following · style_control=true | 1454.7 1449.8–1459.6 | community | LMArena Leaderboard Dataset | |
| Russian (style control)older version arena=text · category=russian · style_control=true | 1451.3 1444.1–1458.6 | community | LMArena Leaderboard Dataset | |
| Polish (style control)older version arena=text · category=polish · style_control=true | 1450.2 1439.8–1460.6 | community | LMArena Leaderboard Dataset | |
| Exclude Ties (style control)older version arena=text · category=exclude_ties · style_control=true | 1448.2 1444.1–1452.4 | community | LMArena Leaderboard Dataset | |
| Overall (style control) arena=text · category=overall · style_control=true | 1447.7 1444.7–1450.7 | community | LMArena Leaderboard Dataset | |
| Industry Business And Management And Financial Operations (style control)older version arena=text · category=industry_business_and_management_and_financial_operations · style_control=true | 1446.5 1440.9–1452.1 | community | LMArena Leaderboard Dataset | |
| Industry Writing And Literature And Language (style control)older version arena=text · category=industry_writing_and_literature_and_language · style_control=true | 1443.9 1438.7–1449.1 | community | LMArena Leaderboard Dataset | |
| German (style control)older version arena=text · category=german · style_control=true | 1442.5 1426.8–1458.1 | community | LMArena Leaderboard Dataset | |
| Creative Writing (style control)older version arena=text · category=creative_writing · style_control=true | 1442.1 1435.8–1448.4 | community | LMArena Leaderboard Dataset | |
| Industry Mathematical (style control)older version arena=text · category=industry_mathematical · style_control=true | 1440.0 1430.4–1449.7 | community | LMArena Leaderboard Dataset | |
| Industry Entertainment And Sports And Media (style control)older version arena=text · category=industry_entertainment_and_sports_and_media · style_control=true | 1433.9 1428.3–1439.4 | community | LMArena Leaderboard Dataset | |
| Non English (style control)older version arena=text · category=non_english · style_control=true | 1433.6 1429.9–1437.3 | community | LMArena Leaderboard Dataset | |
| Math (style control)older version arena=text · category=math · style_control=true | 1432.0 1423.4–1440.7 | community | LMArena Leaderboard Dataset | |
| Japanese (style control)older version arena=text · category=japanese · style_control=true | 1411.6 1393.7–1429.6 | community | LMArena Leaderboard Dataset | |
| Korean (style control)older version arena=text · category=korean · style_control=true | 1402.1 1386.3–1417.8 | community | LMArena Leaderboard Dataset |
knowledge science
| GPQA Diamond | 16K budget | 77.3 % | independent | Epoch AI Benchmarking Hub 2025-08-05 |
| GPQA Diamond | 27K budget | 76.8 % | independent | Epoch AI Benchmarking Hub 2025-08-05 |
| GPQA Diamond implementation=vals-ai | thinking | 76.3 % | independent | Vals AI |
| GPQA Diamond | 73.2 % | independent | Epoch AI Benchmarking Hub 2025-08-05 | |
| GPQA Diamond implementation=vals-ai | 70.0 % | independent | Vals AI | |
| Humanity's Last Exam contamination=Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions. · implementation=scale | thinking | 11.5 % 10.3–12.8 | independent | Scale Labs |
| Humanity's Last Exam contamination=Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions. · implementation=scale | 7.9 % 6.9–9.0 | independent | Scale Labs | |
| MMLU-Pro implementation=vals-ai | thinking | 87.9 % | independent | Vals AI |
| MMLU-Pro implementation=vals-ai | 87.2 % | independent | Vals AI |
long context instruction
| MultiChallenge | thinking | 57.2 % 56.3–58.1 | independent | Scale Labs |
multimodal
| MMMU implementation=vals-ai | thinking | 77.5 % | independent | Vals AI |
| MMMU implementation=vals-ai | 73.7 % | independent | Vals AI | |
| VISTA | thinking | 48.4 % 48.0–48.9 | independent | Scale Labs |
| VISTA | 45.3 % 44.6–45.9 | independent | Scale Labs |
professional
| LegalBench | 83.5 % | independent | Vals AI | |
| MedQA | thinking | 93.6 % | independent | Vals AI |
| MedQA | 92.5 % | independent | Vals AI | |
| TaxEval | thinking | 73.7 % | independent | Vals AI |
| TaxEval | 71.5 % | independent | Vals AI |
reasoning math
| AIME implementation=vals-ai | thinking | 78.2 % | independent | Vals AI |
| AIME implementation=vals-ai | 44.2 % | independent | Vals AI | |
| EnigmaEval contamination=Potential contamination warning: This model was evaluated after the public release of EnigmaEval, allowing model builder access to the prompts and solutions. | thinking | 7.2 % 5.7–8.7 | independent | Scale Labs |
| EnigmaEval contamination=Potential contamination warning: This model was evaluated after the public release of EnigmaEval, allowing model builder access to the prompts and solutions. | 4.8 % 3.6–6.0 | independent | Scale Labs | |
| FrontierMath | 27K budget | 7.2 % | independent | Epoch AI Benchmarking Hub 2025-08-05 |
| FrontierMath | 5.9 % | independent | Epoch AI Benchmarking Hub 2025-08-05 | |
| FrontierMath Tier 4older version | 27K budget | 4.2 % | independent | Epoch AI Benchmarking Hub 2025-08-05 |
| MATH-500 implementation=vals-ai | thinking | 95.4 % | independent | Vals AI |
| MATH-500 implementation=vals-ai | 93.0 % | independent | Vals AI | |
| OTIS Mock AIME 2024–2025 | 27K budget | 68.9 % | independent | Epoch AI Benchmarking Hub 2025-08-05 |
| OTIS Mock AIME 2024–2025 | 16K budget | 64.4 % | independent | Epoch AI Benchmarking Hub 2025-08-05 |
| OTIS Mock AIME 2024–2025 | 40.0 % | independent | Epoch AI Benchmarking Hub 2025-08-05 |
Agent + model results
systems, not bare-model scores
| agent + model Epoch Inspect harness + Claude Opus 4.1 20250805 | SWE-bench Verified | 73.3 % | independent | Epoch AI Benchmarking Hub |
| agent + model terminus-2 + Claude Opus 4.1 20250805 | Terminal-Bench 2.0 | 38.0 % | community | Terminal-Bench Leaderboard |
| agent + model OpenHands + Claude Opus 4.1 20250805 | Terminal-Bench 2.0 | 36.9 % | community | Terminal-Bench Leaderboard |
| agent + model mini-SWE-agent + Claude Opus 4.1 20250805 | Terminal-Bench 2.0 | 35.1 % | community | Terminal-Bench Leaderboard |
| agent + model Claude Code + Claude Opus 4.1 20250805 | Terminal-Bench 2.0 | 34.8 % | community | Terminal-Bench Leaderboard |
These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.
