AI Rankings
LLM Leaderboard: AI Models Ranked by Overall Intelligence
This leaderboard ranks AI language models on the Epoch Capabilities Index, a single score built from many hard benchmarks. Next to each score you can see how people rate the model in Arena's blind votes, what it costs and how much text it can read at once.
Data updated
Top 20 at a glance
Tap a model to see its full page: scores, price and how it compares.
Bars start at 150, not zero, so differences are easier to see.
Full ranking
Showing the top 100 of 270 models.Download the full table (CSV)Data credit: Epoch AI, Arena (formerly LMArena), LiteLLM model price list, used under open licences (details at the foot of the page).
| Rank and model | Epoch Capabilities Index | Likely range | Arena text rating | Price per 1M tokens (in / out) | Context window |
|---|---|---|---|---|---|
1Claude Opus 5.5Anthropic | 168.8 | 165.3 to 173.9 | 1509 | $4.00 / $20.00 | 1M tokens |
2GPT-6 AstraOpenAI | 168.8 | 165.4 to 174.5 | 1477 | $10.00 / $50.00 | 922K tokens |
3GPT-6.1 SolOpenAI | 168.4 | 165.0 to 174.4 | 1480 | $2.00 / $10.00 | 922K tokens |
4Claude Fable 5.1Anthropic | 166.5 | 163.5 to 171.3 | 1503 | $10.00 / $50.00 | 1M tokens |
5Gemini 4 ArgonGoogle | 166.4 | 163.2 to 171.5 | 1517 | $2.00 / $10.00 | 1M tokens |
6Claude Sonnet 5.5Anthropic | 166.2 | 163.1 to 171.3 | 1478 | $2.00 / $10.00 | 1M tokens |
7Claude Opus 5Anthropic | 163.9 | 161.1 to 168.2 | 1491 | $5.00 / $25.00 | 1M tokens |
8GPT-6 SolOpenAI | 163.7 | 160.9 to 167.4 | 1459 | $2.00 / $10.00 | 922K tokens |
9Claude Fable 5Anthropic | 163.4 | 160.5 to 168.2 | 1503 | $10.00 / $50.00 | 1M tokens |
10GPT-5.5 ProOpenAI | 162.9 | 160.0 to 166.6 | Not measured | $30.00 / $180 | 1.1M tokens |
11GPT-5.6 SolOpenAI | 162.8 | 160.2 to 167.1 | 1485 | $4.00 / $20.00 | 922K tokens |
12Muse WonderMeta | 161.8 | 159.4 to 165.4 | 1509 | Not measured | Not measured |
13GPT-5.6 TerraOpenAI | 160.1 | 157.3 to 163.3 | 1467 | $2.00 / $12.00 | 922K tokens |
14GPT-5.5OpenAI | 159.7 | 157.3 to 163.7 | 1482 | $5.00 / $30.00 | 1.1M tokens |
15GPT-5.4 ProOpenAI | 159.6 | 157.1 to 162.7 | Not measured | $30.00 / $180 | 1.1M tokens |
16Claude Haiku 5.5Anthropic | 159.3 | 156.2 to 162.5 | 1425 | $0.10 / $0.50 | 1M tokens |
17Claude Opus 4.8Anthropic | 159.0 | 156.9 to 161.9 | 1482 | $5.00 / $25.00 | 1M tokens |
18Gemini 3.7 FlashGoogle | 158.0 | 155.9 to 161.2 | 1490 | $0.75 / $3.75 | 1M tokens |
19Kimi K3Moonshot AI ยท Open weights | 157.9 | 155.1 to 160.6 | 1487 | $3.00 / $15.00 | 1M tokens |
20Gemini 3.8 FlashGoogle | 157.4 | 155.1 to 161.1 | 1496 | $0.75 / $3.75 | 1M tokens |
21Qwen 3.8 MaxAlibaba | 157.4 | 155.3 to 160.3 | 1483 | $2.00 / $6.00 | 992K tokens |
22GPT-5.4OpenAI | 157.3 | 155.4 to 159.8 | 1475 | $2.50 / $15.00 | 1.1M tokens |
23Muse Spark 1.3Meta | 157.2 | 155.3 to 159.8 | 1496 | $1.25 / $4.25 | 1M tokens |
24GPT-5.3 CodexOpenAI | 157.1 | 153.6 to 161.4 | Not measured | $1.75 / $14.00 | 272K tokens |
25Claude Sonnet 5Anthropic | 156.9 | 154.1 to 159.7 | 1462 | $2.00 / $10.00 | 1M tokens |
26Claude Opus 4.7Anthropic | 156.8 | 154.9 to 159.2 | 1501 | $5.00 / $25.00 | 1M tokens |
27GPT-6 LunaOpenAI | 156.8 | 154.3 to 159.4 | 1442 | $0.10 / $0.50 | 922K tokens |
28Grok 4.6xAI | 156.6 | 154.7 to 159.2 | 1456 | $2.00 / $6.00 | 500K tokens |
29GPT-5.6 LunaOpenAI | 156.5 | 154.2 to 159.3 | 1453 | $0.20 / $1.20 | 922K tokens |
30GPT-5.2 ProOpenAI | 156.3 | 153.9 to 160.0 | Not measured | $21.00 / $168 | 272K tokens |
31GLM-5.3Z.ai ยท Open weights | 156.0 | 154.2 to 158.3 | 1481 | $1.40 / $4.40 | 1M tokens |
32DeepSeek V4.1 FlashDeepSeek ยท Open weights | 155.9 | 153.7 to 158.6 | 1476 | $0.30 / $1.20 | 1M tokens |
33Claude Opus 4.6Anthropic | 155.7 | 153.8 to 158.1 | 1503 | $5.00 / $25.00 | 1M tokens |
34DeepSeek V4 Pro 0813DeepSeek ยท Open weights | 155.6 | 153.7 to 158.0 | 1465 | $1.32 / $3.96 | 1M tokens |
35Gemini 3.1 ProGoogle | 155.3 | 153.3 to 158.1 | 1487 | $2.00 / $12.00 | 1M tokens |
36Muse Spark OpenMeta ยท Open weights | 155.1 | 152.7 to 158.0 | Not measured | $1.25 / $4.25 | 1M tokens |
37Qwen3.8 Max (0902)Alibaba | 155.1 | 153.0 to 157.2 | Not measured | Not measured | Not measured |
38Muse Spark 1.2Meta | 155.1 | 152.8 to 157.6 | 1491 | $1.25 / $4.25 | 1M tokens |
39Gemini 3.5 FlashGoogle | 154.9 | 152.9 to 157.4 | 1477 | $1.50 / $9.00 | 1M tokens |
40DeepSeek V4 Flash 0731DeepSeek ยท Open weights | 154.8 | 152.1 to 156.9 | Not measured | $0.19 / $0.38 | 1M tokens |
41Gemini 3.6 FlashGoogle | 154.5 | 152.8 to 156.6 | 1483 | $0.75 / $3.75 | 1M tokens |
42Grok 4.7xAI | 154.4 | 152.6 to 157.1 | 1441 | $2.00 / $6.00 | 500K tokens |
43Grok 4.5xAI | 154.1 | 152.4 to 156.2 | 1466 | $2.00 / $6.00 | 500K tokens |
44Muse Spark 1.1Meta | 154.1 | 151.7 to 156.2 | 1490 | $1.25 / $4.25 | 1M tokens |
45Qwen3.7-MaxAlibaba | 154.0 | 152.3 to 156.5 | 1474 | $2.50 / $7.50 | 992K tokens |
46GPT-5.2OpenAI | 153.7 | 152.1 to 155.4 | 1437 | $1.75 / $14.00 | 272K tokens |
47GLM-5.3-FlashZ.ai ยท Open weights | 153.1 | 150.8 to 155.5 | 1476 | $0.15 / $0.50 | 1M tokens |
48Gemini 3 ProGoogle | 153.1 | 151.1 to 155.4 | 1486 | $2.00 / $12.00 | 1M tokens |
49Claude Sonnet 4.6Anthropic | 152.8 | 150.2 to 155.1 | 1472 | $3.00 / $15.00 | 1M tokens |
50Grok 4.20xAI | 152.1 | 149.6 to 154.5 | Not measured | $1.25 / $2.50 | 1M tokens |
51Muse SparkMeta | 152.0 | 149.8 to 154.5 | 1489 | Not measured | Not measured |
52GLM-5.2Z.ai ยท Open weights | 151.9 | 150.0 to 154.1 | 1475 | $1.40 / $4.40 | 1M tokens |
53Gemini 3 FlashGoogle | 151.9 | 150.4 to 153.6 | 1473 | $0.50 / $3.00 | 1M tokens |
54Qwen 3.8 27BAlibaba ยท Open weights | 151.7 | 149.8 to 153.7 | 1437 | $0.45 / $3.00 | 262K tokens |
55Kimi K2.6Moonshot AI ยท Open weights | 150.9 | 149.0 to 152.7 | 1461 | $0.95 / $4.00 | 262K tokens |
56GPT-5 ProOpenAI | 150.6 | 148.9 to 154.2 | Not measured | $15.00 / $120 | 400K tokens |
57Inkling-SmallThinking Machines ยท Open weights | 150.3 | 147.2 to 152.1 | 1405 | $0.47 / $1.20 | 1M tokens |
58Claude Opus 4.5Anthropic | 150.1 | 147.8 to 153.0 | 1474 | $5.00 / $25.00 | 200K tokens |
58Kimi K2.7 CodeMoonshot AI ยท Open weights | 150.1 | 148.2 to 151.8 | Not measured | $0.95 / $4.00 | 262K tokens |
60Mistral Large 4Mistral AI | 150.0 | 146.9 to 152.7 | 1434 | $0.68 / $2.09 | 1M tokens |
61GPT-5OpenAI | 150.0 | 150.0 to 150.0 | 1435 | $1.25 / $10.00 | 272K tokens |
62GLM-5.1Z.ai ยท Open weights | 149.8 | 147.7 to 151.9 | 1465 | $1.40 / $4.40 | 200K tokens |
63GPT-5.1OpenAI | 149.8 | 148.4 to 151.7 | 1456 | $1.25 / $10.00 | 272K tokens |
64Qwen 3.6 Max (Preview)Alibaba | 149.4 | 147.6 to 153.2 | 1461 | $1.27 / $7.61 | 262K tokens |
65DeepSeek-V4-ProDeepSeek ยท Open weights | 149.1 | 147.6 to 150.9 | 1458 | $1.32 / $3.96 | 1M tokens |
66Grok 4.3 BetaxAI | 148.8 | 147.2 to 150.3 | Not measured | Not measured | Not measured |
67GPT-5.4 MiniOpenAI | 148.7 | 146.9 to 150.5 | 1447 | $0.75 / $4.50 | 272K tokens |
68InklingThinking Machines ยท Open weights | 148.4 | 146.0 to 150.2 | 1442 | $1.00 / $4.05 | 1M tokens |
69Kimi K2.5Moonshot AI ยท Open weights | 148.1 | 146.6 to 149.5 | 1450 | $0.60 / $3.00 | 262K tokens |
70o3-proOpenAI | 147.3 | 145.6 to 150.1 | Not measured | $20.00 / $80.00 | 200K tokens |
71Qwen 3.6 PlusAlibaba | 147.3 | 145.1 to 149.3 | 1443 | $0.50 / $3.00 | 1M tokens |
72Qwen3.7-PlusAlibaba | 147.0 | 145.3 to 148.8 | 1456 | $0.36 / $1.44 | 1M tokens |
73MiniMax-M3MiniMax ยท Open weights | 146.9 | 142.3 to 150.2 | 1440 | $0.30 / $1.20 | 1M tokens |
74Qwen3.5 397B-A17BAlibaba ยท Open weights | 146.8 | 145.2 to 148.9 | 1442 | $0.60 / $3.60 | 262K tokens |
75Claude Sonnet 4.5Anthropic | 146.7 | 145.0 to 148.3 | 1456 | $3.00 / $15.00 | 200K tokens |
76o3OpenAI | 146.7 | 144.8 to 148.5 | 1432 | $2.00 / $8.00 | 200K tokens |
77MiniMax-M2.5MiniMax ยท Open weights | 146.6 | 142.2 to 147.9 | 1391 | $0.30 / $1.20 | 1M tokens |
78Grok 4xAI | 146.2 | 144.4 to 148.0 | 1411 | $3.00 / $15.00 | 262K tokens |
79Qwen3.6 27BAlibaba ยท Open weights | 146.2 | 144.1 to 147.7 | Not measured | $0.37 / $3.20 | 262K tokens |
80Qwen 3.5 Plus (hosted 397B-A17B)Alibaba | 146.2 | 143.5 to 147.9 | Not measured | Not measured | Not measured |
81DeepSeek-V4-FlashDeepSeek ยท Open weights | 146.1 | 143.8 to 148.0 | 1439 | $0.30 / $1.20 | 1M tokens |
82DeepSeek-V3.2DeepSeek ยท Open weights | 146.1 | 144.3 to 147.4 | 1425 | $0.28 / $0.40 | 164K tokens |
83GPT-5.4 NanoOpenAI | 146.0 | 143.6 to 148.2 | 1401 | $0.20 / $1.25 | 272K tokens |
84Nemotron 3 UltraNVIDIA ยท Open weights | 146.0 | 143.7 to 148.2 | Not measured | Not measured | Not measured |
85Kimi K2 ThinkingMoonshot AI ยท Open weights | 145.8 | 143.2 to 147.5 | Not measured | $0.60 / $2.50 | 262K tokens |
86GLM-5Z.ai ยท Open weights | 145.8 | 143.8 to 147.7 | 1457 | $1.00 / $3.20 | 200K tokens |
87MiniMax-M2.7MiniMax ยท Open weights | 145.8 | 138.4 to 148.2 | 1415 | $0.30 / $1.19 | 197K tokens |
88o4-miniOpenAI | 145.5 | 142.9 to 147.2 | 1391 | $1.10 / $4.40 | 200K tokens |
89GPT-5 miniOpenAI | 145.3 | 143.3 to 147.1 | 1390 | $0.25 / $2.00 | 272K tokens |
90Gemini 2.5 Pro (Jun 2025)Google | 145.1 | 143.4 to 146.7 | 1446 | $1.25 / $10.00 | 1M tokens |
91DeepSeek-V3.2-ExpDeepSeek ยท Open weights | 144.9 | 142.4 to 146.8 | 1425 | $0.27 / $0.41 | 164K tokens |
91Gemini 3.5 Flash-LiteGoogle | 144.9 | 142.4 to 146.3 | 1455 | $0.30 / $2.50 | 1M tokens |
93Gemini 3.1 Flash-LiteGoogle | 144.3 | 142.3 to 145.7 | 1433 | $0.25 / $1.50 | 1M tokens |
94Grok 4 FastxAI | 144.1 | 141.6 to 146.3 | 1419 | Not measured | Not measured |
95Claude Opus 4.1Anthropic | 144.1 | 141.8 to 146.6 | 1450 | $15.00 / $75.00 | 200K tokens |
96Qwen3.7 FlashAlibaba | 144.1 | 141.9 to 145.9 | Not measured | Not measured | Not measured |
97Gemini 2.5 Pro (Mar 2025)Google | 144.0 | 141.7 to 146.8 | Not measured | Not measured | Not measured |
98Qwen3-235B-A22B-Thinking (Jul 2025)Alibaba ยท Open weights | 143.8 | 141.7 to 145.3 | 1400 | $0.30 / $2.90 | 262K tokens |
99Qwen 3.6 35B-A3BAlibaba ยท Open weights | 143.7 | 141.1 to 145.8 | Not measured | $0.20 / $1.10 | 262K tokens |
100GLM-4.7Z.ai ยท Open weights | 143.6 | 141.4 to 145.8 | 1441 | $0.60 / $2.20 | 200K tokens |
Best score over time
Each dot is a model that beat every model released before it. The orange line follows open-weights models only.
- All models: now 168.8 (Claude Opus 5.5)
- Open-weights models: now 157.9 (Kimi K3)
How to read this leaderboard
Higher is better. The index is a relative scale, so only the gaps between models matter, not the number itself. Each score comes with a likely range: when two models' ranges overlap, the difference between them is too small to trust, so treat nearby ranks as a tie. Prices are dollars per million tokens (a token is a small piece of a word).
Questions people ask
What is the Epoch Capabilities Index?
It is a combined score from Epoch AI that merges results from many different benchmarks into one number, so models that were tested on different sets of benchmarks can still be compared. A higher number means stronger results overall.
Why does each score have a range?
Every score is an estimate. The range shows how sure Epoch is: a wide range means less certainty. If the ranges of two models overlap, we cannot say which one is truly ahead.
Does the top model suit my work best?
Not always. A leaderboard compares models on one kind of test. Price, speed, privacy rules, the apps you already use and the type of work you do all matter. Use the rank as a starting point, then try two or three models 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. Ratings come from the open leaderboard dataset published by Arena (arena.ai), licensed CC BY 4.0. We keep the latest published ratings, match models to our other sources and rank them. Arena 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.
- Human preferenceWhich AI do people prefer in blind tests?Leads: Gemini 4 Argon (1517)166 models ranked
- CodingWhich AI model is best for coding?Leads: Gemini 4 Argon (1552)166 models ranked
- Web developmentWhich AI model builds the best websites and apps?Leads: Claude Opus 5.5 (1807)85 models ranked
- Image understandingWhich AI model understands images best?Leads: Claude Fable 5 (1309)85 models ranked
- Open-weights modelsWhich open-source AI model is best?Leads: Kimi K3 (157.9)123 models ranked
- Cheapest strong modelsWhat is the cheapest AI model that is still good?Cheapest: gpt-oss-20b ($0.10)119 models ranked
- Best valueWhich AI models give the most intelligence for the money?Line starts at: gpt-oss-20b ($0.10)7 models ranked
- Context windowWhich AI model can read the most text at once?Leads: GPT-5.4 (1.1M tokens)193 models ranked
- GPQA DiamondWhich AI model scores highest on GPQA Diamond?Leads: Gemini 4 Argon (96.4%)183 models ranked
- Humanity's Last ExamWhich AI model scores highest on Humanity's Last Exam?Leads: GPT-6 Astra (54.8%)41 models ranked
- SWE-bench VerifiedWhich AI model scores highest on SWE-bench Verified?Leads: Claude Opus 4.7 (83.5%)32 models ranked
- Aider PolyglotWhich AI model scores highest on Aider Polyglot?Leads: GPT-5 (88.0%)46 models ranked
- Terminal-BenchWhich AI model scores highest on Terminal-Bench?Leads: GPT-5.3 Codex (64.7%)25 models ranked
- FrontierMath (Tiers 1-3)Which AI model scores highest on FrontierMath (Tiers 1-3)?Leads: GPT-6 Astra (93.7%)87 models ranked
- ARC-AGI-2Which AI model scores highest on ARC-AGI-2?Leads: GPT-6 Astra (95.0%)86 models ranked
- SimpleQA VerifiedWhich AI model scores highest on SimpleQA Verified?Leads: Gemini 4 Argon (76.0%)82 models ranked
- SimpleBenchWhich AI model scores highest on SimpleBench?Leads: Claude Fable 5 (81.9%)80 models ranked
- DeepSWEWhich AI model scores highest on DeepSWE?Leads: GPT-6 Astra (74.1%)26 models ranked
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.
- Arena (formerly LMArena): Human-preference ratings: Arena (arena.ai) leaderboard dataset, CC BY 4.0. We keep the latest published ratings, match them to models from other sources and rank them. Ratings are not changed. Arena 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 ยท Report a mistake