Sign Up

Claude Sonnet 4.5

Anthropic
Announcement
Claude Sonnet 4.5 launch post cover

Cost/1M tokens

input
$3.00
cached
$0.30
output
$15.00
cache write
$3.75
Context
1Mtokens
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 #70

  1. 1511
    Muse WonderSimilar price
  2. 1487
    Gemini 3.1 Pro PreviewSimilar price
  3. 1475
    GPT-5.4Similar price
  4. 1457
    Seed 2.0 ProNearby rank
  5. 1456
    Grok 4.6Nearby rank
  6. 1456
    Claude Sonnet 4.5
  7. 1455
    GPT-5.1Nearby rank
  8. 1455
    Qwen3.7 PlusNearby rank
  9. 1402
    Claude Sonnet 4Predecessor
  10. 1354
    Command ASimilar price

+54 vs Claude Sonnet 4 (predecessor) · Ahead of 1 of 4 similarly priced models

Arena WebDev

Best Arena rating per model · Higher is better

Arena #109

  1. 1658
    Kimi K3Similar price
  2. 1649
    Muse WonderSimilar price
  3. 1462
    GPT-5.4Similar price
  4. 1446
    Gemini 3.1 Pro PreviewSimilar price
  5. 1398
    GPT-5.4 MiniNearby rank
  6. 1397
    GPT-5.1Nearby rank
  7. 1393
    Claude Sonnet 4.5
  8. 1389
    Claude Opus 4.1Nearby rank
  9. 1386
    MiniMax M2.5Nearby 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

Arena #35

  1. 1471
    GPT-5.4Similar price
  2. 1466
    Grok 4.5Similar price
  3. 1459
    Gemini 3.1 Pro PreviewSimilar price
  4. 1450
    Kimi K2.6Nearby rank
  5. 1450
    Grok 4.6Nearby rank
  6. 1448
    Claude Sonnet 4.5
  7. 1446
    Qwen3.7 PlusNearby rank
  8. 1438
    Kimi K2.5Nearby rank
  9. 1425
    GPT-5.2Similar price

Ahead of 1 of 4 similarly priced models

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

#53 of 85

  1. 77.1%
    Gemini 3.1 Pro PreviewSimilar price
  2. 74.0%
    GPT-5.4Similar price
  3. 67.1%
    Grok 4.6Similar price
  4. 60.4%
    Kimi K3Similar price
  5. 18.3%
    GPT-5 ProNearby rank
  6. 17.6%
    GPT-5.1Nearby rank
  7. 13.6%
    Claude Sonnet 4.5
  8. 11.8%
    Kimi K2.5Nearby rank
  9. 10.3%
    Gemini 3.5 Flash LiteNearby rank
  10. 5.9%
    Claude Sonnet 4Predecessor

+7.7 pts vs Claude Sonnet 4 (predecessor) · Ahead of 0 of 4 similarly priced models

FrontierMath-2025-02-28-Private

Best score per model · Higher is better

#28 of 58

  1. 47.6%
    GPT-5.4Similar price
  2. 40.7%
    GPT-5.2Similar price
  3. 36.9%
    Gemini 3.1 Pro PreviewSimilar price
  4. 18.7%
    o3Nearby rank
  5. 16.4%
    GLM 5Nearby rank
  6. 15.2%
    Claude Sonnet 4.5
  7. 14.1%
    Gemini 2.5 ProNearby rank
  8. 10.3%
    Gemini 2.5 Pro Preview 06-05Nearby rank
  9. 4.1%
    Claude Sonnet 4Predecessor
  10. 0.3%
    GPT-4o (2024-08-06)Similar price

+11.1 pts vs Claude Sonnet 4 (predecessor) · Ahead of 1 of 4 similarly priced models

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

Best score per model · Higher is better

#18 of 46

  1. 27.1%
    GPT-5.4Similar price
  2. 18.8%
    GPT-5.2Similar price
  3. 16.7%
    Gemini 3.1 Pro PreviewSimilar price
  4. 4.2%
    Claude Opus 4.1Nearby rank
  5. 4.2%
    Claude Opus 4.5Nearby rank
  6. 4.2%
    Claude Sonnet 4.5
  7. 4.2%
    Gemini 2.5 FlashNearby rank
  8. 4.2%
    Gemini 2.5 ProNearby rank
  9. 2.1%
    o3Similar price
  10. 0.0%
    Claude Sonnet 4Predecessor

+4.2 pts vs Claude Sonnet 4 (predecessor) · Ahead of 1 of 4 similarly priced models

FrontierMath-Tier-4-v2-Private

Best score per model · Higher is better

#64 of 72

  1. 49.0%
    GPT-5.4Similar price
  2. 39.0%
    Kimi K3Similar price
  3. 34.1%
    Qwen3.8 Max (0902)Similar price
  4. 26.8%
    Gemini 3.1 Pro PreviewSimilar price
  5. 4.9%
    InklingNearby rank
  6. 2.4%
    Claude Opus 4.1Nearby rank
  7. 2.4%
    Claude Sonnet 4.5
  8. 2.4%
    DeepSeek V4 Pro 0423Nearby rank
  9. 2.4%
    GPT-5 NanoNearby rank

Ahead of 0 of 4 similarly priced models

FrontierMath-Tiers-1-3-v2-Private

Best score per model · Higher is better

#74 of 87

  1. 78.6%
    GPT-5.4Similar price
  2. 72.2%
    Kimi K3Similar price
  3. 66.0%
    Grok 4.6Similar price
  4. 59.6%
    Gemini 3.1 Pro PreviewSimilar price
  5. 26.0%
    Gemini 3.5 Flash LiteNearby rank
  6. 24.6%
    Gemini 2.5 ProNearby rank
  7. 23.9%
    Claude Sonnet 4.5
  8. 22.5%
    Qwen3.6 FlashNearby rank
  9. 20.4%
    Qwen3.6 35B A3BNearby rank

Ahead of 0 of 4 similarly priced models

GPQA diamond

Best score per model · Higher is better

#81 of 161

  1. 94.4%
    Gemini 3.1 Pro PreviewSimilar price
  2. 94.0%
    Grok 4.6Similar price
  3. 93.3%
    GPT-5.4Similar price
  4. 93.1%
    Kimi K3Similar price
  5. 83.3%
    GLM 4.7Nearby rank
  6. 82.5%
    GPT-5.5 InstantNearby rank
  7. 82.3%
    Claude Sonnet 4.5
  8. 82.3%
    Qwen3.5-FlashNearby rank
  9. 82.3%
    Qwen3.7 FlashNearby rank
  10. 79.2%
    Claude Sonnet 4Predecessor

+3.1 pts vs Claude Sonnet 4 (predecessor) · Ahead of 0 of 4 similarly priced models

Humanity's Last Exam

Best score per model · Higher is better

#27 of 34

  1. 46.4%
    Gemini 3.1 Pro PreviewSimilar price
  2. 36.2%
    GPT-5.4Similar price
  3. 27.8%
    GPT-5.2Similar price
  4. 8.0%
    Claude Sonnet 3.7Nearby rank
  5. 7.9%
    Claude Opus 4.1Nearby rank
  6. 7.5%
    Claude Sonnet 4.5
  7. 6.7%
    Claude Opus 4Nearby rank
  8. 5.7%
    Llama 4 MaverickNearby rank
  9. 5.5%
    Claude Sonnet 4Predecessor
  10. 2.7%
    GPT-4o (2024-11-20)Similar price

+2.0 pts vs Claude Sonnet 4 (predecessor) · Ahead of 1 of 4 similarly priced models

OSWorld

Best score per model · Higher is better

#5 of 10

  1. 72.1%
    Claude Sonnet 4.6Similar price
  2. 66.3%
    Claude Opus 4.5Similar price
  3. 63.3%
    Kimi K2.5Nearby rank
  4. 62.9%
    Claude Sonnet 4.5
  5. 43.9%
    Claude Sonnet 4Predecessor
  6. 35.8%
    Claude Sonnet 3.7Nearby rank
  7. 23.0%
    o3Similar price

+19.0 pts vs Claude Sonnet 4 (predecessor) · Ahead of 1 of 3 similarly priced models

OTIS Mock AIME 2024-2025

Best score per model · Higher is better

#87 of 158

  1. 100.0%
    Qwen3.8 Max (0902)Similar price
  2. 97.8%
    GPT-5.4Similar price
  3. 97.2%
    Kimi K3Similar price
  4. 95.6%
    Gemini 3.1 Pro PreviewSimilar price
  5. 80.0%
    Gemini 3.1 Flash LiteNearby rank
  6. 80.0%
    GLM 5Nearby rank
  7. 77.8%
    Claude Sonnet 4.5
  8. 73.3%
    Qwen3 MaxNearby rank
  9. 73.3%
    Gemma 4 31BNearby rank
  10. 71.1%
    Claude Sonnet 4Predecessor

+6.7 pts vs Claude Sonnet 4 (predecessor) · Ahead of 0 of 4 similarly priced models

SimpleQA Verified

Best score per model · Higher is better

#66 of 87

  1. 73.5%
    Gemini 3.1 Pro PreviewSimilar price
  2. 56.0%
    Grok 4.7Similar price
  3. 50.6%
    Kimi K3Similar price
  4. 45.1%
    GPT-5.4Similar price
  5. 32.2%
    GLM 4.7Nearby rank
  6. 31.1%
    GPT-4.1Nearby rank
  7. 30.7%
    Claude Sonnet 4.5
  8. 30.2%
    Grok 4.20 (Reasoning)Nearby rank
  9. 29.4%
    GPT-5.4 MiniNearby rank

Ahead of 0 of 4 similarly priced models

SWE-bench Verified

Best score per model · Higher is better

#25 of 35

  1. 76.9%
    GPT-5.4Similar price
  2. 75.6%
    Gemini 3.1 Pro Preview Custom ToolsSimilar price
  3. 74.8%
    GPT-5.3-CodexSimilar price
  4. 73.8%
    GPT-5.2Similar price
  5. 72.9%
    Gemini 3 Pro PreviewNearby rank
  6. 72.1%
    GLM 5Nearby rank
  7. 71.3%
    Claude Sonnet 4.5
  8. 70.7%
    Claude Opus 4Nearby rank
  9. 65.9%
    GPT-5.1Nearby rank
  10. 61.0%
    Claude Sonnet 3.7Earlier version

+10.3 pts vs Claude Sonnet 3.7 (earlier version) · Ahead of 0 of 4 similarly priced models

Terminal Bench

Best score per model · Higher is better

#21 of 49

  1. 81.8%
    GPT-5.4Similar price
  2. 80.2%
    Gemini 3.1 Pro PreviewSimilar price
  3. 78.4%
    GPT-5.3-CodexSimilar price
  4. 66.5%
    GPT-5.2-CodexSimilar price
  5. 49.6%
    GPT-5Nearby rank
  6. 47.6%
    GPT-5.1Nearby rank
  7. 46.5%
    Claude Sonnet 4.5
  8. 45.1%
    MiniMax M2.7Nearby rank
  9. 44.3%
    GPT-5-CodexNearby rank

Ahead of 0 of 4 similarly priced models

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.