Skip to main content

AI Rankings

GPQA Diamond Leaderboard: AI Models Ranked by Score

This page ranks AI models by their GPQA Diamond score, using Epoch AI's open benchmark data.

Data updated

What GPQA Diamond measures

Very hard multiple-choice science questions in biology, physics and chemistry, written by experts and designed so a quick web search does not give the answer.

Scores are percentages and higher is better. Results come from Epoch AI's open benchmark data; we do not run the test ourselves.

Top 20 at a glance

Tap a model to see its full page: scores, price and how it compares.
Top 20 models on GPQA Diamond scoreRanked list of 20 items. 1. Gemini 4 Argon, 96.4%; 2. GPT-6 Astra, 95.8%; 3. Claude Sonnet 5.5, 95.6%.
  1. Gemini 4 Argon96.4%Released 30 Sep 2026
  2. GPT-6 Astra95.8%Released 3 Sep 2026
  3. Claude Sonnet 5.595.6%Released 28 Sep 2026
  4. Gemini 3.8 Flash95.4%Released 2 Sep 2026
  5. GPT-6.1 Sol95.4%Released 29 Sep 2026
  6. Gemini 3.7 Flash94.8%Released 13 Aug 2026
  7. GPT-5.4 Pro94.6%Released 5 Mar 2026
  8. Gemini 3.1 Pro94.4%Released 19 Feb 2026
  9. GPT-6 Sol94.3%Released 22 Sep 2026
  10. Gemini 3.6 Flash94.1%Released 21 Jul 2026
  11. GPT-5.594.0%Released 23 Apr 2026
  12. Grok 4.694.0%Released 12 Aug 2026
  13. GPT-5.5 Pro93.9%Released 23 Apr 2026
  14. Claude Opus 593.9%Released 24 Jul 2026
  15. GPT-5.6 Sol93.5%Released 9 Jul 2026
  16. Grok 4.593.4%Released 8 Jul 2026
  17. GPT-5.6 Terra93.3%Released 9 Jul 2026
  18. GPT-5.493.3%Released 5 Mar 2026
  19. Kimi K393.1%Open weightsReleased 16 Jul 2026
  20. Gemini 3.5 Flash92.8%Released 19 May 2026

Full ranking

Showing the top 100 of 183 models.Download the full table (CSV)Data credit: Epoch AI, LiteLLM model price list, used under open licences (details at the foot of the page).
GPQA Diamond Leaderboard: AI Models Ranked by Score: 100 of 183 models
Rank and modelGPQA Diamond scoreEpoch indexPrice per 1M tokens (in / out)Released
1Gemini 4 ArgonGoogle
96.4%166.4$2.00 / $10.002026-09-30
2GPT-6 AstraOpenAI
95.8%168.8$10.00 / $50.002026-09-03
3Claude Sonnet 5.5Anthropic
95.6%166.2$2.00 / $10.002026-09-28
4Gemini 3.8 FlashGoogle
95.4%157.4$0.75 / $3.752026-09-02
4GPT-6.1 SolOpenAI
95.4%168.4$2.00 / $10.002026-09-29
6Gemini 3.7 FlashGoogle
94.8%158.0$0.75 / $3.752026-08-13
7GPT-5.4 ProOpenAI
94.6%159.6$30.00 / $1802026-03-05
8Gemini 3.1 ProGoogle
94.4%155.3$2.00 / $12.002026-02-19
9GPT-6 SolOpenAI
94.3%163.7$2.00 / $10.002026-09-22
10Gemini 3.6 FlashGoogle
94.1%154.5$0.75 / $3.752026-07-21
11GPT-5.5OpenAI
94.0%159.7$5.00 / $30.002026-04-23
11Grok 4.6xAI
94.0%156.6$2.00 / $6.002026-08-12
13GPT-5.5 ProOpenAI
93.9%162.9$30.00 / $1802026-04-23
14Claude Opus 5Anthropic
93.9%163.9$5.00 / $25.002026-07-24
15GPT-5.6 SolOpenAI
93.5%162.8$4.00 / $20.002026-07-09
16Grok 4.5xAI
93.4%154.1$2.00 / $6.002026-07-08
17GPT-5.6 TerraOpenAI
93.3%160.1$2.00 / $12.002026-07-09
18GPT-5.4OpenAI
93.3%157.3$2.50 / $15.002026-03-05
19Kimi K3Moonshot AI · Open weights
93.1%157.9$3.00 / $15.002026-07-16
20Gemini 3.5 FlashGoogle
92.8%154.9$1.50 / $9.002026-05-19
21Grok 4.7xAI
92.7%154.4$2.00 / $6.002026-09-21
22Qwen 3.8 MaxAlibaba
92.7%157.4$2.00 / $6.002026-08-02
23Gemini 3 ProGoogle
92.6%153.1$2.00 / $12.002025-11-18
24Qwen3.8 Max (0902)Alibaba
92.3%155.1Not measured2026-09-01
25GLM-5.2Z.ai · Open weights
91.9%151.9$1.40 / $4.402026-06-16
26DeepSeek V4 Pro 0813DeepSeek · Open weights
91.7%155.6$1.32 / $3.962026-08-13
27GPT-5.6 LunaOpenAI
91.6%156.5$0.20 / $1.202026-07-09
28GPT-5.2OpenAI
91.4%153.7$1.75 / $14.002025-12-11
29Claude Opus 4.8Anthropic
91.0%159.0$5.00 / $25.002026-05-28
29DeepSeek V4 Flash 0731DeepSeek · Open weights
91.0%154.8$0.19 / $0.382026-07-31
31DeepSeek-V4-ProDeepSeek · Open weights
90.9%149.1$1.32 / $3.962026-04-24
31GLM-5.3Z.ai · Open weights
90.9%156.0$1.40 / $4.402026-08-14
31MiniMax-M3MiniMax · Open weights
90.9%146.9$0.30 / $1.202026-06-01
31Qwen3.7-MaxAlibaba
90.9%154.0$2.50 / $7.502026-05-19
35Kimi K2.6Moonshot AI · Open weights
90.8%150.9$0.95 / $4.002026-04-20
36Claude Opus 5.5Anthropic
90.6%168.8$4.00 / $20.002026-09-22
37Claude Opus 4.6Anthropic
90.5%155.7$5.00 / $25.002026-02-05
37Claude Sonnet 5Anthropic
90.5%156.9$2.00 / $10.002026-06-30
39GPT-6 LunaOpenAI
90.5%156.8$0.10 / $0.502026-09-22
40Claude Opus 4.7Anthropic
90.2%156.8$5.00 / $25.002026-04-16
40GLM-5.3-FlashZ.ai · Open weights
90.2%153.1$0.15 / $0.502026-08-20
42GLM-5.1Z.ai · Open weights
89.9%149.8$1.40 / $4.402026-04-07
43DeepSeek V4.1 FlashDeepSeek · Open weights
89.8%155.9$0.30 / $1.202026-09-09
44Muse SparkMeta
89.8%152.0Not measured2026-04-08
45Claude Haiku 5.5Anthropic
89.7%159.3$0.10 / $0.502026-10-07
46Gemini 3 FlashGoogle
89.4%151.9$0.50 / $3.002025-12-17
47Grok 4.20xAI
89.3%152.1$1.25 / $2.502026-02-17
48Grok 4.3 BetaxAI
88.8%148.8Not measured2026-04-17
49Muse Spark OpenMeta · Open weights
88.6%155.1$1.25 / $4.252026-10-15
50Inkling-SmallThinking Machines · Open weights
88.5%150.3$0.47 / $1.202026-07-15
51Qwen 3.6 PlusAlibaba
88.4%147.3$0.50 / $3.002026-03-31
52InklingThinking Machines · Open weights
88.3%148.4$1.00 / $4.052026-07-15
53Mistral Large 4Mistral AI
88.1%150.0$0.68 / $2.092026-10-06
54Kimi K2.7 CodeMoonshot AI · Open weights
87.9%150.1$0.95 / $4.002026-06-12
54Qwen3.7-PlusAlibaba
87.9%147.0$0.36 / $1.442026-06-02
56GLM-5Z.ai · Open weights
87.8%145.8$1.00 / $3.202026-02-11
57GPT-5.1OpenAI
87.6%149.8$1.25 / $10.002025-11-13
58Kimi K2.5Moonshot AI · Open weights
87.6%148.1$0.60 / $3.002026-01-27
59Claude Sonnet 4.6Anthropic
87.4%152.8$3.00 / $15.002026-02-17
59Qwen 3.6 Max (Preview)Alibaba
87.4%149.4$1.27 / $7.612026-04-20
61Grok 4xAI
87.0%146.2$3.00 / $15.002025-07-09
62GPT-5.4 MiniOpenAI
86.9%148.7$0.75 / $4.502026-03-17
63Qwen3.5 397B-A17BAlibaba · Open weights
86.4%146.8$0.60 / $3.602026-02-13
64GPT-5OpenAI
86.2%150.0$1.25 / $10.002025-08-07
65Claude Opus 4.5Anthropic
86.0%150.1$5.00 / $25.002025-11-24
66Claude Fable 5Anthropic
85.9%163.4$10.00 / $50.002026-06-09
66Qwen3.6 27BAlibaba · Open weights
85.9%146.2$0.37 / $3.202026-04-22
68Nemotron 3 UltraNVIDIA · Open weights
85.3%146.0Not measured2026-06-04
69Gemini 2.5 Pro (Jun 2025)Google
85.3%145.1$1.25 / $10.002025-06-05
70Qwen 3.5 Plus (hosted 397B-A17B)Alibaba
84.8%146.2Not measured2026-02-16
70Qwen 3.6 35B-A3BAlibaba · Open weights
84.8%143.7$0.20 / $1.102026-04-14
72Kimi K2 ThinkingMoonshot AI · Open weights
84.2%145.8$0.60 / $2.502025-11-06
73Gemini 2.5 Pro (Mar 2025)Google
83.8%144.0Not measured2025-03-31
74Qwen3.5-35B-A3BAlibaba · Open weights
83.5%142.5$0.25 / $1.252026-02-24
75DeepSeek-V3.2DeepSeek · Open weights
83.4%146.1$0.28 / $0.402025-12-01
76Gemini 3.5 Flash-LiteGoogle
83.3%144.9$0.30 / $2.502026-07-21
76GLM-4.7Z.ai · Open weights
83.3%143.6$0.60 / $2.202025-12-22
76Qwen 3.6 FlashAlibaba
83.3%142.9Not measured2026-04-27
79GPT-5.5 InstantOpenAI
82.5%142.4Not measured2026-05-05
80Claude Sonnet 4.5Anthropic
82.3%146.7$3.00 / $15.002025-09-29
80Qwen 3.5 Flash (hosted 35B-A3B)Alibaba
82.3%143.5Not measured2026-02-25
80Qwen3.7 FlashAlibaba
82.3%144.1Not measured2026-07-27
83Gemini 3.1 Flash-LiteGoogle
81.8%144.3$0.25 / $1.502026-03-03
83o3OpenAI
81.8%146.7$2.00 / $8.002025-04-16
85Qwen3-235B-A22B-Thinking (Jul 2025)Alibaba · Open weights
80.0%143.8$0.30 / $2.902025-07-25
86Claude 3.7 SonnetAnthropic
79.7%141.0$3.00 / $15.002025-02-24
87o4-miniOpenAI
79.6%145.5$1.10 / $4.402025-04-16
88Claude Sonnet 4Anthropic
79.2%141.6$3.00 / $15.002025-05-22
89Qwen3.5-9BAlibaba · Open weights
79.0%139.3$0.17 / $0.152026-02-24
90GPT-5.4 NanoOpenAI
78.5%146.0$0.20 / $1.252026-03-17
91Claude Opus 4.1Anthropic
77.3%144.1$15.00 / $75.002025-08-05
92o3-miniOpenAI
77.0%140.2$1.10 / $4.402025-01-31
93o1OpenAI
76.8%141.7$15.00 / $60.002024-12-17
94DeepSeek-R1 (May 2025)DeepSeek · Open weights
76.3%141.1$0.75 / $2.452025-05-28
95Claude Opus 4Anthropic
76.3%142.5$15.00 / $75.002025-05-22
95Grok-3 minixAI
76.3%140.2$0.28 / $0.892025-06-24
97Gemma 4 31B ITGoogle · Open weights
75.8%143.0$0.14 / $0.402026-04-02
97gpt-oss-120bOpenAI · Open weights
75.8%139.8$0.15 / $0.602025-08-05
97Grok 3xAI
75.8%138.2$3.00 / $15.002025-04-09
100GPT-5 miniOpenAI
75.0%145.3$0.25 / $2.002025-08-07

How to read this leaderboard

Scores are percentages and higher is better. For each model we show the best result Epoch lists across its effort settings, the same rule for every model. A gap of a point or two is usually within the normal noise of a test, so treat close scores as a tie. Newer models can be missing when nobody has run the test on them yet, so check the release dates. A benchmark measures one kind of skill: use it together with the overall leaderboard rather than on its own.

Questions people ask

What does GPQA Diamond measure?

Very hard multiple-choice science questions in biology, physics and chemistry, written by experts and designed so a quick web search does not give the answer.

Who ran these tests?

The results come from Epoch AI's open benchmark data. It combines runs by Epoch with results published by other groups, and Epoch's own pages say which is which. We do not run tests ourselves.

Does a high score mean the model is best for me?

Not by itself. GPQA Diamond checks one kind of ability. A model can lead here and still be slower, costlier or a poor fit for your work. Compare it on the overall leaderboard and try it on your own tasks.

Why is a model I know missing?

A model appears only when an open source has published results for it. New models can take days or weeks to be measured, and models with no published score are left out rather than guessed.

Where does the data come from?

Scores come from Epoch AI's open benchmark data (licensed CC BY 4.0). We match models across sources, rank them and show them in charts, without changing the scores. Epoch AI does not endorse this site. Prices and context windows come from the open-source LiteLLM price list (MIT licence), which records each model maker's own price page. Prices change often, so confirm on the maker's site before you buy.

How often is this leaderboard updated?

We check our open data sources every day and show the date of the data at the top of the page. A model only moves when its source publishes new results, so a number can stay the same for weeks.

More leaderboards

Different questions need different tests. Pick the one that matches your work.

Where this data comes from

  • Epoch AI: Benchmark data: Epoch AI, 'Capabilities & benchmarking', epoch.ai/benchmarks (CC BY 4.0). We match models across sources, rank them and show them in charts. Scores are not changed. Epoch AI does not endorse this site.
  • LiteLLM model price list: Prices and context windows: LiteLLM model price list (github.com/BerriAI/litellm, MIT licence). We use the model maker's own price entry when there is one. Prices change often, so always confirm on the maker's site before you buy.

Newest data on this page: 26 Oct 2026. Prices are list prices and change often, so confirm on the maker's site before you buy. Sponsors never change a score or an order. How we calculate this · Report a mistake

We use optional cookies for analytics and referral tracking. Learn more about our privacy policy