Sign Up

Claude Opus 4

Anthropic
Announcement
Claude Opus 4 launch post cover
Context
200Ktokens
Released
1y ago

Benchmarks

DeepSWE is the primary coding benchmark where an exact model evaluation exists. Arena leaderboards and other sourced evaluations appear below; missing scores are not zero.

How it ranks

Against its predecessor, similarly priced models and the models ranked around it.

DeepSWE 1.1

Best pass@1 per model · Higher is better

Not evaluated in DeepSWE 1.1

Arena Text

Best Arena rating per model · Higher is better

Arena #123

  1. 1427
    Mistral Medium 3.5Nearby rank
  2. 1426
    Nemotron 3 UltraNearby rank
  3. 1426
    Claude Opus 4
  4. 1425
    DeepSeek V3.2 ExpNearby rank
  5. 1425
    DeepSeek V3.2Nearby rank
  6. 1421
    R1 0528Similar release date
  7. 1402
    Qwen3 235B-A22BSimilar release date
  8. 1388
    Claude Sonnet 3.7Predecessor
  9. 1387
    Mistral Medium 3Similar release date
  10. 1347
    Qwen3 32BSimilar release date

+38 vs Claude Sonnet 3.7 (predecessor) · Ahead of 3 of 4 models released around the same time

Arena WebDev

Best Arena rating per model · Higher is better

Not ranked on Arena WebDev yet

Arena Vision

Best Arena rating per model · Higher is better

Arena #88

  1. 1247
    Gemini 2.5 ProSimilar release date
  2. 1216
    o3Similar release date
  3. 1214
    GPT-4.1Nearby rank
  4. 1211
    GPT-5Nearby rank
  5. 1208
    Claude Opus 4
  6. 1208
    Inkling SmallNearby rank
  7. 1206
    Claude Sonnet 4Nearby rank
  8. 1193
    Claude Sonnet 3.7Predecessor
  9. 1155
    Mistral Medium 3Similar release date
  10. 1145
    Llama 4 MaverickSimilar release date

Level with Claude Sonnet 3.7 (predecessor, overlapping 95% CIs) · Ahead of 2 of 4 models released around the same time

Arena Search

Best Arena rating per model · Higher is better

Not ranked on Arena Search yet

Arena Document

Best Arena rating per model · Higher is better

Not ranked on Arena Document yet

Agent Arena

Best IPS score per model · Higher is better

Not ranked on Agent Arena yet

ARC-AGI-2

Best score per model · Higher is better

#58 of 85

  1. 10.3%
    Gemini 3.5 Flash LiteNearby rank
  2. 9.9%
    GPT-5Nearby rank
  3. 8.6%
    Claude Opus 4
  4. 6.5%
    o3Nearby rank
  5. 5.9%
    Claude Sonnet 4Nearby rank
  6. 4.9%
    o3 ProSimilar release date
  7. 2.5%
    Gemini 2.5 Flash Preview 05-20Similar release date
  8. 1.1%
    R1 0528Similar release date
  9. 0.9%
    Claude Sonnet 3.7Predecessor
  10. 0.0%
    Magistral SmallSimilar release date

+7.7 pts vs Claude Sonnet 3.7 (predecessor) · Ahead of 4 of 4 models released around the same time

FrontierMath-2025-02-28-Private

Best score per model · Higher is better

#42 of 58

  1. 18.7%
    o3Similar release date
  2. 10.3%
    Gemini 2.5 Pro Preview 06-05Similar release date
  3. 5.5%
    GPT-4.1Nearby rank
  4. 4.8%
    Gemini 2.5 FlashNearby rank
  5. 4.5%
    Claude Opus 4
  6. 4.5%
    GPT-4.1 MiniNearby rank
  7. 4.1%
    Claude Sonnet 3.7Predecessor
  8. 0.3%
    Mistral Medium 3Similar release date
  9. 0.0%
    Llama 4 ScoutSimilar release date

+0.3 pts vs Claude Sonnet 3.7 (predecessor) · Ahead of 2 of 4 models released around the same time

FrontierMath-Tier-4-2025-07-01-Private

Best score per model · Higher is better

#18 of 46

  1. 6.3%
    GPT-5.4 NanoNearby rank
  2. 4.2%
    Qwen3.6 Max PreviewNearby rank
  3. 4.2%
    Claude Opus 4
  4. 4.2%
    Claude Opus 4.1Nearby rank
  5. 4.2%
    Claude Opus 4.5Nearby rank
  6. 2.1%
    GLM 4.6Similar release date
  7. 2.1%
    Gemini 2.5 Pro Preview 06-05Similar release date
  8. 2.1%
    o3Similar release date
  9. 0.0%
    Claude Sonnet 3.5 v2Earlier version
  10. 0.0%
    Qwen3 235B A22B Thinking 2507Similar release date

+4.2 pts vs Claude Sonnet 3.5 v2 (earlier version) · Ahead of 4 of 4 models released around the same time

GPQA diamond

Best score per model · Higher is better

#97 of 161

  1. 84.8%
    Gemini 2.5 Pro Preview 06-05Similar release date
  2. 79.7%
    Claude Sonnet 3.7Predecessor
  3. 77.3%
    Claude Opus 4.1Nearby rank
  4. 76.3%
    R1 0528Similar release date
  5. 76.3%
    Claude Opus 4
  6. 75.8%
    Gemma 4 31BNearby rank
  7. 75.8%
    gpt-oss-120bNearby rank
  8. 63.8%
    Qwen3 14BSimilar release date
  9. 59.5%
    Mistral Medium 3Similar release date

-3.5 pts vs Claude Sonnet 3.7 (predecessor) · Ahead of 2 of 4 models released around the same time

Humanity's Last Exam

Best score per model · Higher is better

#28 of 34

  1. 20.3%
    o3Similar release date
  2. 11.0%
    Gemini 2.5 Flash Preview 05-20Similar release date
  3. 8.0%
    Claude Sonnet 3.7Predecessor
  4. 7.9%
    Claude Opus 4.1Nearby rank
  5. 7.5%
    Claude Sonnet 4.5Nearby rank
  6. 6.7%
    Claude Opus 4
  7. 5.7%
    Llama 4 MaverickSimilar release date
  8. 5.5%
    Claude Sonnet 4Nearby rank
  9. 4.5%
    Mistral Medium 3Similar release date

-1.4 pts vs Claude Sonnet 3.7 (predecessor) · Ahead of 2 of 4 models released around the same time

OTIS Mock AIME 2024-2025

Best score per model · Higher is better

#109 of 158

  1. 70.8%
    Gemini 2.5 Flash Preview 05-20Similar release date
  2. 66.4%
    R1 0528Similar release date
  3. 66.4%
    Qwen3 14BSimilar release date
  4. 65.3%
    gpt-oss-20bNearby rank
  5. 64.4%
    Claude Opus 4
  6. 62.8%
    Qwen3 30B A3BNearby rank
  7. 62.2%
    Qwen3 30B A3B Instruct 2507Nearby rank
  8. 57.8%
    Claude Sonnet 3.7Predecessor
  9. 32.2%
    Mistral Medium 3Similar release date

+6.7 pts vs Claude Sonnet 3.7 (predecessor) · Ahead of 1 of 4 models released around the same time

SWE-bench Verified

Best score per model · Higher is better

#27 of 35

  1. 73.6%
    GPT-5Similar release date
  2. 72.1%
    GLM 5Nearby rank
  3. 71.3%
    Claude Sonnet 4.5Nearby rank
  4. 70.7%
    Claude Opus 4
  5. 65.9%
    GPT-5.1Nearby rank
  6. 64.7%
    GPT-5 MiniNearby rank
  7. 62.3%
    o3Similar release date
  8. 61.0%
    Claude Sonnet 3.7Predecessor
  9. 57.6%
    Gemini 2.5 ProSimilar release date
  10. 48.5%
    GPT-4.1Similar release date

+9.7 pts vs Claude Sonnet 3.7 (predecessor) · Ahead of 3 of 4 models released around the same time

Not evaluated in DeepSWE 1.1 as of 2026-09-22. Results are only shown for exact model versions in the official dataset. Official leaderboard.

All published scores

  • Arena TextArena #1231426

    +38 vs Claude Sonnet 3.7 (predecessor)

    • default142616K, thinking95% CI 1422 to 1431 · 35,852 votes
    • default141495% CI 1410 to 1419 · 42,967 votes
    Source:Arena Text leaderboard, 2026-11-09
  • Arena VisionArena #881208

    Level with Claude Sonnet 3.7 (predecessor, overlapping 95% CIs)

    • default120816K, thinking95% CI 1192 to 1224 · 1,287 votes
    • default118995% CI 1176 to 1202 · 2,166 votes
    Source:Arena Vision leaderboard, 2026-11-09
  • ARC-AGI-2#58 of 858.6%

    +7.7 pts vs Claude Sonnet 3.7 (predecessor)

    • default8.6%16K
    • default4.5%8K
    • default1.3%
    • default0.0%1K
    Source:External evaluation via Epoch AIOriginal-source terms apply
  • FrontierMath-2025-02-28-Private#42 of 584.5%

    +0.3 pts vs Claude Sonnet 3.7 (predecessor)

    • default4.5%
    • default4.1%27K
    Source:Epoch AI evaluation
  • FrontierMath-Tier-4-2025-07-01-Private#18 of 464.2%

    +4.2 pts vs Claude Sonnet 3.5 v2 (earlier version)

    • default4.2%27K
    • default0.0%
    Source:Epoch AI evaluation
  • GPQA diamond#97 of 16176.3%

    -3.5 pts vs Claude Sonnet 3.7 (predecessor)

    • default76.3%16K
    • default69.2%
    Source:Epoch AI evaluation
  • Humanity's Last Exam#28 of 346.7%

    -1.4 pts vs Claude Sonnet 3.7 (predecessor)

    • unknown6.7%
    Source:External evaluation via Epoch AIOriginal-source terms apply
  • OTIS Mock AIME 2024-2025#109 of 15864.4%

    +6.7 pts vs Claude Sonnet 3.7 (predecessor)

    • default64.4%27K
    • default60.0%16K
    • default42.2%
    Source:Epoch AI evaluation
  • SWE-bench Verified#27 of 3570.7%

    +9.7 pts vs Claude Sonnet 3.7 (predecessor)

    • default70.7%
    Source:Epoch AI evaluation

Benchmark compilation: Epoch AI, Capabilities & benchmarking (CC BY attribution). Scores are converted to percentages and model names matched to the Vertical Tabs catalog. Individual evaluations are credited to their original sources. External results may carry separate licences and evaluation configurations; do not interpret scores from different harnesses as directly comparable.

Arena ratings: Arena (lmarena-ai/leaderboard-dataset), licensed CC BY 4.0. Ratings come from pairwise human votes and are only comparable within one leaderboard; the interval is Arena's 95% confidence interval. Model names are matched to the Vertical Tabs catalog.

To give you the best experience, we use cookies for personalization, insights, and performance. Learn more.