Sign Up

Gemini 2.5 Pro

Google DeepMind
Announcement
Gemini 2.5 Pro launch post cover

Cost/1M tokens

input
$1.25
cached
$0.125
output
$10.00
cache write
$0.375
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 #89

  1. 1472
    Claude Sonnet 5.5Similar price
  2. 1466
    Grok 4.5Similar price
  3. 1455
    GPT-5.1Similar price
  4. 1447
    GPT-5.4 MiniNearby rank
  5. 1446
    ERNIE 5.0Nearby rank
  6. 1446
    Gemini 2.5 Pro
  7. 1443
    InklingNearby rank
  8. 1443
    Qwen3.6 PlusNearby rank
  9. 1427
    Mistral Medium 3.5Similar price
  10. 1360
    Gemini 2.0 FlashPredecessor

+85 vs Gemini 2.0 Flash (predecessor) · Ahead of 1 of 4 similarly priced models

Arena WebDev

Best Arena rating per model · Higher is better

Arena #149

  1. 1676
    Qwen3.8 Max (0902)Similar price
  2. 1641
    Grok 4.7Similar price
  3. 1417
    GPT-5Similar price
  4. 1263
    Mistral Medium 3.5Similar price
  5. 1242
    Grok 4.1 Fast (Reasoning)Nearby rank
  6. 1233
    Mistral Large 3 2512Nearby rank
  7. 1228
    Gemini 2.5 Pro
  8. 1196
    Devstral 2 2512Nearby rank
  9. 1164
    x-AI/Grok-Code-Fast 1Nearby rank

Ahead of 0 of 4 similarly priced models

Arena Vision

Best Arena rating per model · Higher is better

Arena #57

  1. 1280
    Grok 4.5Similar price
  2. 1263
    Claude Sonnet 5.5Similar price
  3. 1251
    GPT-5.1Similar price
  4. 1249
    Grok 4.7Nearby rank
  5. 1248
    MiMo-V2.6-FlashNearby rank
  6. 1247
    Gemini 2.5 Pro
  7. 1247
    MiMo-V2.6-ProNearby rank
  8. 1246
    Qwen3.5 397B A17BNearby rank
  9. 1199
    Mistral Medium 3.5Similar price
  10. 1172
    Gemini 2.0 FlashPredecessor

+75 vs Gemini 2.0 Flash (predecessor) · Ahead of 1 of 4 similarly priced models

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 #41

  1. 1480
    Muse Spark 1.1Similar price
  2. 1469
    Claude Sonnet 5Similar price
  3. 1466
    Grok 4.5Similar price
  4. 1438
    Kimi K2.5Nearby rank
  5. 1435
    MiniMax M3Nearby rank
  6. 1433
    Gemini 2.5 Pro
  7. 1425
    GPT-5.2Nearby rank
  8. 1419
    Claude Haiku 4.5Nearby rank
  9. 1416
    GPT-5.1Similar 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

#62 of 85

  1. 83.9%
    Claude Sonnet 5.5Similar price
  2. 67.1%
    Grok 4.6Similar price
  3. 60.4%
    Kimi K3Similar price
  4. 17.6%
    GPT-5.1Similar price
  5. 5.9%
    Claude Sonnet 4Nearby rank
  6. 5.7%
    GPT-5.4 NanoNearby rank
  7. 4.9%
    Gemini 2.5 Pro
  8. 4.9%
    MiniMax M2.5Nearby rank
  9. 4.9%
    o3 ProNearby rank

Ahead of 0 of 4 similarly priced models

FrontierMath-2025-02-28-Private

Best score per model · Higher is better

#30 of 58

  1. 33.4%
    GLM 5.1Similar price
  2. 32.4%
    GPT-5Similar price
  3. 31.0%
    GPT-5.1Similar price
  4. 16.4%
    GLM 5Nearby rank
  5. 15.2%
    Claude Sonnet 4.5Nearby rank
  6. 14.1%
    Gemini 2.5 Pro
  7. 10.3%
    Gemini 2.5 Pro Preview 06-05Nearby rank
  8. 10.3%
    Qwen3.6 FlashNearby rank
  9. 5.9%
    Claude Haiku 4.5Similar price

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. 12.5%
    GPT-5Similar price
  2. 12.5%
    GPT-5.1Similar price
  3. 12.5%
    GLM 5.1Similar price
  4. 4.2%
    Claude Sonnet 4.5Nearby rank
  5. 4.2%
    Gemini 2.5 FlashNearby rank
  6. 4.2%
    Gemini 2.5 Pro
  7. 4.2%
    Gemini 3 Flash PreviewNearby rank
  8. 2.1%
    GLM 4.6Nearby rank
  9. 2.1%
    Claude Haiku 4.5Similar price

Ahead of 1 of 4 similarly priced models

FrontierMath-Tier-4-v2-Private

Best score per model · Higher is better

#71 of 72

  1. 80.5%
    Claude Sonnet 5.5Similar price
  2. 34.1%
    Qwen3.8 Max (0902)Similar price
  3. 31.7%
    Grok 4.6Similar price
  4. 22.0%
    GPT-5Similar price
  5. 2.4%
    GPT-5 NanoNearby rank
  6. 2.4%
    GPT-5.5 InstantNearby rank
  7. 0.0%
    Gemini 2.5 Pro
  8. 0.0%
    Gemini 3.5 Flash LiteNearby rank

Ahead of 0 of 4 similarly priced models

FrontierMath-Tiers-1-3-v2-Private

Best score per model · Higher is better

#73 of 87

  1. 88.8%
    Claude Sonnet 5.5Similar price
  2. 66.0%
    Grok 4.6Similar price
  3. 65.6%
    Qwen3.8 Max (0902)Similar price
  4. 55.4%
    GPT-5Similar price
  5. 26.3%
    GPT-5.5 InstantNearby rank
  6. 26.0%
    Gemini 3.5 Flash LiteNearby rank
  7. 24.6%
    Gemini 2.5 Pro
  8. 23.9%
    Claude Sonnet 4.5Nearby rank
  9. 22.5%
    Qwen3.6 FlashNearby rank

Ahead of 0 of 4 similarly priced models

GPQA diamond

Best score per model · Higher is better

#70 of 161

  1. 94.0%
    Grok 4.6Similar price
  2. 92.3%
    Qwen3.8 Max (0902)Similar price
  3. 87.6%
    GPT-5.1Similar price
  4. 85.9%
    Qwen3.6 27BNearby rank
  5. 85.4%
    Nemotron 3 UltraNearby rank
  6. 85.3%
    Gemini 2.5 Pro
  7. 84.8%
    Qwen3.5 PlusNearby rank
  8. 84.8%
    Gemini 2.5 Pro Preview 06-05Nearby rank
  9. 49.0%
    Mistral Large 2407Similar price

Ahead of 1 of 4 similarly priced models

OTIS Mock AIME 2024-2025

Best score per model · Higher is better

#76 of 158

  1. 100.0%
    Qwen3.8 Max (0902)Similar price
  2. 99.2%
    Grok 4.6Similar price
  3. 91.4%
    GPT-5Similar price
  4. 86.1%
    Claude Opus 4.5Nearby rank
  5. 85.8%
    Claude Sonnet 4.6Nearby rank
  6. 84.7%
    Gemini 2.5 Pro
  7. 84.4%
    GPT-5.5Nearby rank
  8. 84.4%
    o3Nearby rank
  9. 8.5%
    Mistral Large 2407Similar price

Ahead of 1 of 4 similarly priced models

SWE-bench Verified

Best score per model · Higher is better

#33 of 35

  1. 77.3%
    Qwen3.7 MaxSimilar price
  2. 75.2%
    Claude Sonnet 4.6Similar price
  3. 74.2%
    GLM 5.1Similar price
  4. 73.6%
    GPT-5Similar price
  5. 61.0%
    Claude Sonnet 3.7Nearby rank
  6. 57.9%
    Qwen3.6 PlusNearby rank
  7. 57.6%
    Gemini 2.5 Pro
  8. 48.5%
    GPT-4.1Nearby rank
  9. 31.0%
    GPT-4o (2024-11-20)Nearby rank

Ahead of 0 of 4 similarly priced models

Terminal Bench

Best score per model · Higher is better

#37 of 49

  1. 60.4%
    GPT-5.1-Codex-MaxSimilar price
  2. 57.8%
    GPT-5.1-CodexSimilar price
  3. 49.6%
    GPT-5Similar price
  4. 35.5%
    Claude Haiku 4.5Similar price
  5. 34.8%
    GPT-5 MiniNearby rank
  6. 33.4%
    GLM 4.7Nearby rank
  7. 32.6%
    Gemini 2.5 Pro
  8. 30.0%
    MiniMax-M2Nearby rank
  9. 27.8%
    Kimi K2 InstructNearby 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.