Sign Up

Claude Opus 5

Anthropic
Announcement
Claude Opus 5 launch post cover

Cost/1M tokens

input
$5.00
cached
$0.50
output
$25.00
cache write
$6.25
Context
1Mtokens
Released
3mo 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

#3 of 28

  1. 74.1%
    GPT-6 Astra
  2. 73.8%
    Gemini 3.8 Flash
  3. 73.6%
    Claude Opus 5
  4. 72.7%
    GPT-5.6 Sol
  5. 69.9%
    Claude Fable 5
  6. 69.6%
    GPT-5.6 Terra
  7. 69.0%
    GLM 5.3
  8. 68.5%
    Kimi K3
  9. 67.5%
    Grok 4.6
  10. 67.2%
    GPT-5.6 Luna

Arena Text

Best Arena rating per model · Higher is better

Arena #15

  1. 1511
    Muse WonderSimilar price
  2. 1493
    Muse Spark 1.2Nearby rank
  3. 1491
    Muse Spark 1.1Nearby rank
  4. 1490
    Claude Opus 5
  5. 1489
    Gemini 3.7 FlashNearby rank
  6. 1487
    Kimi K3Nearby rank
  7. 1485
    GPT-5.6 SolSimilar price
  8. 1482
    Claude Opus 4.8Predecessor
  9. 1481
    GPT-5.5Similar price
  10. 1346
    GPT-4o (2024-05-13)Similar price

+8 vs Claude Opus 4.8 (predecessor) · Ahead of 2 of 4 similarly priced models

Arena WebDev

Best Arena rating per model · Higher is better

Arena #10

  1. 1742
    Claude Fable 5.1Nearby rank
  2. 1723
    Muse Wonder FullNearby rank
  3. 1691
    Claude Opus 5
  4. 1688
    Kimi K3.1 ProNearby rank
  5. 1682
    Gemini 4 ArgonNearby rank
  6. 1649
    Muse WonderSimilar price
  7. 1615
    GPT-5.6 SolSimilar price
  8. 1555
    Claude Opus 4.8Predecessor
  9. 1512
    GPT-5.5Similar price
  10. 1462
    GPT-5.4Similar price

+136 vs Claude Opus 4.8 (predecessor) · Ahead of 4 of 4 similarly priced models

Arena Vision

Best Arena rating per model · Higher is better

Arena #19

  1. 1288
    GPT-5.6 SolSimilar price
  2. 1286
    Claude Opus 4.8Predecessor
  3. 1285
    Claude Opus 5
  4. 1285
    GPT-5.5Similar price
  5. 1283
    GPT-6 AstraNearby rank
  6. 1282
    GPT-5.4Similar price
  7. 1162
    GPT-4o (2024-05-13)Similar price

Level with Claude Opus 4.8 (predecessor, overlapping 95% CIs) · Ahead of 1 of 4 similarly priced models

Arena Search

Best Arena rating per model · Higher is better

Arena #10

  1. 1255
    GPT-5.6 SolSimilar price
  2. 1245
    GPT-6 AstraSimilar price
  3. 1227
    Qwen3.8 MaxNearby rank
  4. 1227
    Claude Fable 5.1Nearby rank
  5. 1213
    Claude Opus 5
  6. 1213
    Claude Opus 5.5Nearby rank
  7. 1211
    GPT-6 LunaNearby rank
  8. 1210
    Claude Opus 4.7Similar price
  9. 1194
    Claude Opus 4.8Predecessor
  10. 1142
    Sonar ProSimilar price

+19 vs Claude Opus 4.8 (predecessor) · Ahead of 1 of 4 similarly priced models

Arena Document

Best Arena rating per model · Higher is better

Arena #8

  1. 1494
    Claude Opus 4.7Nearby rank
  2. 1486
    GPT-5.6 SolSimilar price
  3. 1486
    Claude Opus 5
  4. 1485
    Claude Opus 4.8Predecessor
  5. 1481
    GPT-5.5Similar price
  6. 1474
    GPT-6 AstraSimilar price
  7. 1471
    GPT-5.4Similar price

Level with Claude Opus 4.8 (predecessor, overlapping 95% CIs) · Ahead of 1 of 4 similarly priced models

Agent Arena

Best IPS score per model · Higher is better

Arena #12

  1. +0.195
    Claude Opus 5.5Similar price
  2. +0.145
    GPT-6 AstraSimilar price
  3. +0.090
    Muse Wonder FullNearby rank
  4. +0.073
    Claude Fable 5Nearby rank
  5. +0.072
    Claude Opus 5
  6. +0.069
    GPT-6 SolNearby rank
  7. +0.053
    Hy4Nearby rank
  8. +0.046
    GPT-5.6 SolSimilar price
  9. +0.018
    Muse WonderSimilar price

Ahead of 2 of 4 similarly priced models

ARC-AGI-2

Best score per model · Higher is better

#12 of 85

  1. 95.0%
    GPT-6 AstraSimilar price
  2. 92.5%
    GPT-5.6 SolSimilar price
  3. 89.2%
    Claude Fable 5Nearby rank
  4. 89.2%
    Gemini 3.8 FlashNearby rank
  5. 88.3%
    Claude Opus 5
  6. 85.0%
    GPT-5.5Similar price
  7. 84.6%
    Gemini 3.7 FlashNearby rank
  8. 74.0%
    GPT-5.4Similar price
  9. 72.1%
    Claude Opus 4.8Predecessor

+16.3 pts vs Claude Opus 4.8 (predecessor) · Ahead of 2 of 4 similarly priced models

FrontierMath-Tier-4-v2-Private

Best score per model · Higher is better

#15 of 72

  1. 97.6%
    GPT-6 AstraSimilar price
  2. 82.9%
    GPT-5.6 SolSimilar price
  3. 78.0%
    Gemini 4 ArgonNearby rank
  4. 78.0%
    GPT-5.5 ProNearby rank
  5. 73.2%
    Claude Opus 5
  6. 72.5%
    GPT-5.5Similar price
  7. 70.7%
    GPT-5.6 TerraNearby rank
  8. 56.1%
    Claude Opus 4.8Predecessor
  9. 49.0%
    GPT-5.4Similar price

+17.1 pts vs Claude Opus 4.8 (predecessor) · Ahead of 2 of 4 similarly priced models

FrontierMath-Tiers-1-3-v2-Private

Best score per model · Higher is better

#15 of 87

  1. 93.7%
    GPT-6 AstraSimilar price
  2. 89.1%
    GPT-5.6 SolSimilar price
  3. 86.0%
    Gemini 4 ArgonNearby rank
  4. 86.0%
    GPT-5.6 TerraNearby rank
  5. 85.6%
    Claude Opus 5
  6. 85.3%
    GPT-5.5Similar price
  7. 82.5%
    GPT-5.4 ProNearby rank
  8. 80.0%
    Claude Opus 4.8Predecessor
  9. 78.6%
    GPT-5.4Similar price

+5.6 pts vs Claude Opus 4.8 (predecessor) · Ahead of 2 of 4 similarly priced models

GPQA diamond

Best score per model · Higher is better

#14 of 161

  1. 94.0%
    Grok 4.6Nearby rank
  2. 93.9%
    Claude Fable 5.2Nearby rank
  3. 93.9%
    Claude Opus 5
  4. 93.5%
    GPT-5.6 SolSimilar price
  5. 93.4%
    Grok 4.5Nearby rank
  6. 93.3%
    GPT-5.4Similar price
  7. 91.0%
    Claude Opus 4.8Predecessor
  8. 90.7%
    GPT-5.5Similar price
  9. 48.9%
    GPT-4o (2024-05-13)Similar price

+2.8 pts vs Claude Opus 4.8 (predecessor) · Ahead of 4 of 4 similarly priced models

OSWorld 2.0

Best score per model · Higher is better

#1 of 9

  1. 31.4%
    Claude Opus 5
  2. 27.3%
    GPT-5.6 SolSimilar price
  3. 18.5%
    Claude Opus 4.8Predecessor
  4. 13.9%
    Claude Opus 4.7Similar price
  5. 13.0%
    GPT-5.5Similar price
  6. 9.3%
    Claude Sonnet 4.6Similar price

+12.9 pts vs Claude Opus 4.8 (predecessor) · Ahead of 4 of 4 similarly priced models

OTIS Mock AIME 2024-2025

Best score per model · Higher is better

#19 of 158

  1. 100.0%
    GPT-5.6 SolSimilar price
  2. 99.2%
    Grok 4.6Nearby rank
  3. 98.9%
    Claude Haiku 5.5Nearby rank
  4. 98.9%
    Claude Opus 5
  5. 98.9%
    Gemini 3.8 FlashNearby rank
  6. 98.9%
    GPT-6 LunaNearby rank
  7. 98.3%
    Claude Opus 4.8Predecessor
  8. 97.8%
    GPT-5.4Similar price
  9. 84.4%
    GPT-5.5Similar price
  10. 6.3%
    GPT-4o (2024-05-13)Similar price

+0.6 pts vs Claude Opus 4.8 (predecessor) · Ahead of 3 of 4 similarly priced models

SimpleQA Verified

Best score per model · Higher is better

#19 of 87

  1. 75.6%
    GPT-6 AstraSimilar price
  2. 69.7%
    GPT-5.6 SolSimilar price
  3. 63.0%
    GPT-5.5Similar price
  4. 60.7%
    GPT-6 SolNearby rank
  5. 60.3%
    Muse Spark 1.2Nearby rank
  6. 59.9%
    Claude Opus 5
  7. 57.8%
    Muse Spark 1.1Nearby rank
  8. 57.3%
    GPT-6.1 TerraNearby rank
  9. 53.0%
    Claude Opus 4.8Predecessor
  10. 45.1%
    GPT-5.4Similar price

+6.9 pts vs Claude Opus 4.8 (predecessor) · Ahead of 1 of 4 similarly priced models

DeepSWE 1.1

Primary coding-agent benchmark · 113 tasks · Data generated 2026-09-22

Official leaderboard
73.6%
pass@1 · highest published configuration
mini-swe-agent · max reasoning
pass@4
88.5%
Average cost / attempt
$11.84
Benchmark runs
4

Score vs cost by reasoning effort

Average cost per attempt (USD) against pass rate

All published configurations

Harness / reasoningpass@1pass@4Cost / attempt
mini-swe-agent · max73.6%88.5%$11.84
mini-swe-agent · xhigh73.2%85.8%$9.07
mini-swe-agent · high72.8%87.6%$6.08
mini-swe-agent · medium68.9%89.4%$3.29
mini-swe-agent · low58.1%85.0%$1.66

pass@1 is the success rate across scored attempts; pass@4 is the share of tasks solved in at least one of four attempts. Costs are reported by DeepSWE for each configuration. These results are not directly comparable to evaluations using different harnesses. Official source JSON.

Reasoning effort

Arena Text

Score by reasoning effort · Higher is better

Arena WebDev

Score by reasoning effort · Higher is better

ARC-AGI-2

Score by reasoning effort · Higher is better

GPQA diamond

Score by reasoning effort · Higher is better

OSWorld 2.0

Score by reasoning effort · Higher is better

OTIS Mock AIME 2024-2025

Score by reasoning effort · Higher is better

All published scores

  • Arena TextArena #151490

    +8 vs Claude Opus 4.8 (predecessor)

    • high149095% CI 1486 to 1493 · 73,697 votes
    • max148995% CI 1484 to 1494 · 35,710 votes
    Source:Arena Text leaderboard, 2026-11-09
  • Arena WebDevArena #101691

    +136 vs Claude Opus 4.8 (predecessor)

    • high165795% CI 1651 to 1662 · 26,940 votes
    • max169195% CI 1685 to 1697 · 19,561 votes
    Source:Arena WebDev leaderboard, 2026-11-10
  • Arena VisionArena #191285

    Level with Claude Opus 4.8 (predecessor, overlapping 95% CIs)

    • high128595% CI 1279 to 1292 · 19,345 votes
    Source:Arena Vision leaderboard, 2026-11-09
  • Arena SearchArena #101213

    +19 vs Claude Opus 4.8 (predecessor)

    • high121395% CI 1206 to 1220 · 34,000 votes
    Source:Arena Search leaderboard, 2026-11-09
  • Arena DocumentArena #81486

    Level with Claude Opus 4.8 (predecessor, overlapping 95% CIs)

    • high148695% CI 1480 to 1492 · 16,189 votes
    Source:Arena Document leaderboard, 2026-11-09
  • Agent ArenaArena #12+0.072
    • max+0.07295% CI +0.059 to +0.085 · 4,693,134 observations
    Source:Agent Arena leaderboard, 2026-11-11
  • ARC-AGI-2#12 of 8588.3%

    +16.3 pts vs Claude Opus 4.8 (predecessor)

    • high88.3%
    • max88.3%
    Source:External evaluation via Epoch AIOriginal-source terms apply
  • FrontierMath-Tier-4-v2-Private#15 of 7273.2%

    +17.1 pts vs Claude Opus 4.8 (predecessor)

    • max73.2%
    Source:Epoch AI evaluation
  • FrontierMath-Tiers-1-3-v2-Private#15 of 8785.6%

    +5.6 pts vs Claude Opus 4.8 (predecessor)

    • max85.6%
    Source:Epoch AI evaluation
  • GPQA diamond#14 of 16193.9%

    +2.8 pts vs Claude Opus 4.8 (predecessor)

    • low87.9%
    • max93.9%
    • default92.9%
    Source:Epoch AI evaluation
  • OSWorld 2.0#1 of 931.4%

    +12.9 pts vs Claude Opus 4.8 (predecessor)

    • low22.3%
    • medium25.3%
    • high29.0%
    • xhigh30.2%
    • max31.4%
    Source:https://osworld-v2.xlang.ai/Original-source terms apply
  • OTIS Mock AIME 2024-2025#19 of 15898.9%

    +0.6 pts vs Claude Opus 4.8 (predecessor)

    • low93.3%
    • max98.9%
    • default97.8%
    Source:Epoch AI evaluation
  • SimpleQA Verified#19 of 8759.9%

    +6.9 pts vs Claude Opus 4.8 (predecessor)

    • max59.9%
    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.