Sign Up

Claude Opus 4.1

Anthropic
Announcement
Claude Opus 4.1 launch post cover

Cost/1M tokens

input
$15.00
cached
$1.50
output
$75.00
cache write
$18.75
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 #80

  1. 1504
    Claude Fable 5Similar price
  2. 1503
    Claude Fable 5.1Similar price
  3. 1474
    GPT-6 AstraSimilar price
  4. 1453
    GPT-5.6 LunaNearby rank
  5. 1452
    Gemma 4 31BNearby rank
  6. 1450
    Claude Opus 4.1
  7. 1450
    Kimi K2.5Nearby rank
  8. 1448
    Step 5 PreviewNearby rank
  9. 1426
    Claude Opus 4Predecessor

+24 vs Claude Opus 4 (predecessor) · Ahead of 0 of 3 similarly priced models

Arena WebDev

Best Arena rating per model · Higher is better

Arena #110

  1. 1836
    Claude Fable 5.2Similar price
  2. 1764
    GPT-6 AstraSimilar price
  3. 1742
    Claude Fable 5.1Similar price
  4. 1625
    Claude Fable 5Similar price
  5. 1397
    GPT-5.1Nearby rank
  6. 1393
    Claude Sonnet 4.5Nearby rank
  7. 1389
    Claude Opus 4.1
  8. 1386
    MiniMax M2.5Nearby rank
  9. 1370
    Gemma 4 31BNearby rank

Ahead of 0 of 4 similarly priced models

Arena Vision

Best Arena rating per model · Higher is better

Not ranked on Arena Vision yet

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

FrontierMath-2025-02-28-Private

Best score per model · Higher is better

#35 of 58

  1. 8.5%
    Qwen3 235B A22B Thinking 2507Nearby rank
  2. 8.3%
    GPT-5 NanoNearby rank
  3. 7.2%
    Claude Opus 4.1
  4. 6.2%
    Qwen3.5-FlashNearby rank
  5. 5.9%
    Claude Haiku 4.5Nearby rank
  6. 4.5%
    Claude Opus 4Predecessor

+2.8 pts vs Claude Opus 4 (predecessor)

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

Best score per model · Higher is better

#18 of 46

  1. 14.6%
    GPT-5 ProSimilar price
  2. 4.2%
    Qwen3.6 Max PreviewNearby rank
  3. 4.2%
    Claude Opus 4Predecessor
  4. 4.2%
    Claude Opus 4.1
  5. 4.2%
    Claude Opus 4.5Nearby rank
  6. 4.2%
    Claude Sonnet 4.5Nearby rank

Level with Claude Opus 4 (predecessor) · Ahead of 0 of 1 similarly priced model

FrontierMath-Tier-4-v2-Private

Best score per model · Higher is better

#64 of 72

  1. 97.6%
    GPT-6 AstraSimilar price
  2. 90.2%
    Claude Fable 5Similar price
  3. 46.0%
    GPT-5.2 ProSimilar price
  4. 19.5%
    GPT-5 ProSimilar price
  5. 4.9%
    Claude Opus 4.5Nearby rank
  6. 4.9%
    InklingNearby rank
  7. 2.4%
    Claude Opus 4.1
  8. 2.4%
    Claude Sonnet 4.5Nearby rank
  9. 2.4%
    DeepSeek V4 Pro 0423Nearby rank

Ahead of 0 of 4 similarly priced models

FrontierMath-Tiers-1-3-v2-Private

Best score per model · Higher is better

#82 of 87

  1. 93.7%
    GPT-6 AstraSimilar price
  2. 90.2%
    Claude Fable 5.1Similar price
  3. 74.0%
    GPT-5.2 ProSimilar price
  4. 55.8%
    GPT-5 ProSimilar price
  5. 18.9%
    Qwen3 MaxNearby rank
  6. 18.2%
    Qwen3.5-FlashNearby rank
  7. 12.6%
    Claude Opus 4.1
  8. 6.7%
    GPT-4.1 MiniNearby rank
  9. 6.0%
    GPT-4.1Nearby rank

Ahead of 0 of 4 similarly priced models

GPQA diamond

Best score per model · Higher is better

#94 of 161

  1. 95.8%
    GPT-6 AstraSimilar price
  2. 93.9%
    Claude Fable 5.2Similar price
  3. 85.9%
    Claude Fable 5Similar price
  4. 78.5%
    GPT-5.4 NanoNearby rank
  5. 77.6%
    Seed 1.6Nearby rank
  6. 77.3%
    Claude Opus 4.1
  7. 76.3%
    R1 0528Nearby rank
  8. 76.3%
    Claude Opus 4Predecessor

+1.0 pts vs Claude Opus 4 (predecessor) · Ahead of 0 of 3 similarly priced models

Humanity's Last Exam

Best score per model · Higher is better

#26 of 34

  1. 54.8%
    GPT-6 AstraSimilar price
  2. 46.5%
    Claude Fable 5.1Similar price
  3. 31.6%
    GPT-5 ProSimilar price
  4. 8.1%
    GLM 4.5 AirNearby rank
  5. 8.0%
    Claude Sonnet 3.7Nearby rank
  6. 7.9%
    Claude Opus 4.1
  7. 7.5%
    Claude Sonnet 4.5Nearby rank
  8. 6.7%
    Claude Opus 4Predecessor

+1.2 pts vs Claude Opus 4 (predecessor) · Ahead of 0 of 3 similarly priced models

OTIS Mock AIME 2024-2025

Best score per model · Higher is better

#99 of 158

  1. 100.0%
    Claude Fable 5Similar price
  2. 100.0%
    Claude Fable 5.1Similar price
  3. 100.0%
    GPT-6 AstraSimilar price
  4. 99.7%
    Claude Fable 5.2Similar price
  5. 70.0%
    Qwen3.5-35B-A3BNearby rank
  6. 69.4%
    Seed 1.6Nearby rank
  7. 68.9%
    Claude Opus 4.1
  8. 68.1%
    GPT-5.5 InstantNearby rank
  9. 67.5%
    Seed-OSS-36B-InstructNearby rank
  10. 64.4%
    Claude Opus 4Predecessor

+4.4 pts vs Claude Opus 4 (predecessor) · Ahead of 0 of 4 similarly priced models

SWE-bench Verified

Best score per model · Higher is better

#21 of 35

  1. 73.8%
    GPT-5.2Nearby rank
  2. 73.6%
    GPT-5Nearby rank
  3. 73.3%
    Claude Opus 4.1
  4. 72.9%
    Gemini 3 Pro PreviewNearby rank
  5. 72.1%
    GLM 5Nearby rank
  6. 70.7%
    Claude Opus 4Predecessor

+2.7 pts vs Claude Opus 4 (predecessor)

Terminal Bench

Best score per model · Higher is better

#28 of 49

  1. 42.7%
    MiniMax M2.5Nearby rank
  2. 39.6%
    DeepSeek V3.2Nearby rank
  3. 38.0%
    Claude Opus 4.1
  4. 36.6%
    MiniMax-M2.1Nearby rank
  5. 35.7%
    Kimi K2 ThinkingNearby rank

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

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.