Model leaderboard
201 current models: intelligence scores, list price, context and the next shutdown date.
| Lifecycle | Intelligence | Price · $ per 1M tokens | Specs | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| # | Status | |||||||||
| 1 | Gemini 4 ArgonGoogle | Active | None set | 1525 | 96.4% | 1678 | $2.00 | $10.00 | 1M | Oct 2026 |
| 2 | Claude Opus 5.5Anthropic | Active | ≥ 11 moClaude API | 1507 | 90.6% | 1813 | $4.00 | $20.00 | 1M | Sep 2026 |
| 3 | Claude Opus 4.6Anthropic | Active | 4 moAzure AI Foundry | 1504 | 90.5% | 1547 | $5.00 | $25.00 | 1M | Feb 2026 |
| 4 | Claude Fable 5Anthropic | Active | ≥ 8 moGoogle Vertex AI | 1504 | 85.9% | 1625 | $10.00 | $50.00 | 1M | Jun 2026 |
| 5 | Claude Opus 4.7Anthropic | Active | 6 moAzure AI Foundry | 1501 | 90.2% | 1558 | $5.00 | $25.00 | 1M | Apr 2026 |
| 6 | Claude Fable 5.1Anthropic | Active | ≥ 5 moGoogle Vertex AI | 1501 | – | 1744 | $10.00 | $50.00 | 1M | Sep 2026 |
| 7 | Gemini 3.8 FlashGoogle | Active | None set | 1497 | 95.4% | 1584 | $0.75 | $3.75 | 1M | Sep 2026 |
| 8 | Claude Opus 5Anthropic | Active | ≥ 3 moGoogle Vertex AI | 1490 | 93.9% | 1691 | $5.00 | $25.00 | 1M | Jul 2026 |
| 9 | Kimi K3Moonshot AI | Active | None set | 1488 | 93.1% | 1654 | $3.30 | $16.50 | 1M | – |
| 10 | Gemini 3.1 Pro PreviewGoogle | Preview | None set | 1487 | 94.4% | 1447 | $2.00 | $12.00 | 1M | Feb 2026 |
| 11 | Gemini 3.7 FlashGoogle | Legacy | 3 moGoogle Vertex AI | 1487 | 94.8% | 1592 | – | – | 1M | Aug 2026 |
| 12 | GPT-5.6 SolOpenAI | Active | ≥ 9 moAmazon Bedrock | 1485 | 93.5% | – | $4.00 | $20.00 | 1.1M | Jul 2026 |
| 13 | GPT-6.1 SolOpenAI | Active | None set | 1484 | 95.4% | 1755 | $2.00 | $10.00 | 1.1M | Sep 2026 |
| 14 | Gemini 3.6 FlashGoogle | Active | 36 daysGoogle Vertex AI | 1482 | 94.1% | 1538 | $0.75 | $3.75 | 1M | Jul 2026 |
| 15 | GPT-5.5OpenAI | Active | ≥ 8 moAmazon Bedrock | 1482 | 90.7% | – | $5.00 | $30.00 | 1.1M | Apr 2026 |
| 16 | Claude Opus 4.8Anthropic | Active | ≥ 7 moClaude API | 1481 | 91.0% | 1556 | $5.00 | $25.00 | 1M | May 2026 |
| 17 | GLM-5.3Z.ai (GLM) | Active | None set | 1478 | 90.9% | 1622 | – | – | 1M | Aug 2026 |
| 18 | Gemini 3.5 FlashGoogle | Active | ≥ 7 moGoogle Vertex AI | 1478 | 92.8% | 1499 | – | – | 1M | May 2026 |
| 19 | Claude Sonnet 5.5Anthropic | Active | ≥ 11 moClaude API | 1476 | 95.6% | 1774 | $2.00 | $10.00 | 1M | Sep 2026 |
| 20 | GLM-5.2Z.ai (GLM) | Active | 17 daysMistral API | 1475 | 91.9% | 1603 | $1.40 | $4.40 | 1M | Jun 2026 |
| 21 | GPT-5.4OpenAI | Active | ≥ 8 moAmazon Bedrock | 1475 | 93.3% | 1400 | $2.50 | $15.00 | 1.1M | Mar 2026 |
| 22 | GPT-6 AstraOpenAI | Active | ≥ 11 moAmazon Bedrock | 1475 | 95.8% | 1786 | $10.00 | $50.00 | 1.1M | Sep 2026 |
| 23 | GLM-5.3-FlashZ.ai (GLM) | Active | None set | 1475 | 90.2% | 1609 | $0.15 | $0.50 | – | Aug 2026 |
| 24 | DeepSeek-V4.1-FlashDeepSeek | Active | None set | 1475 | 89.8% | 1619 | $0.30 | $1.20 | 1M | – |
| 25 | Claude Opus 4.5Anthropic | Active | ≥ 41 daysClaude API | 1474 | 86.0% | 1494 | $5.00 | $25.00 | 200K | Nov 2025 |
| 26 | Claude Sonnet 4.6Anthropic | Active | 4 moAzure AI Foundry | 1472 | 87.4% | 1522 | $3.00 | $15.00 | 1M | Feb 2026 |
| 27 | GPT-5.6 TerraOpenAI | Active | ≥ 9 moAmazon Bedrock | 1466 | 93.3% | – | $2.00 | $12.00 | 1.1M | Jul 2026 |
| 28 | Grok 4.5xAI | Active | None set | 1466 | 93.4% | 1553 | $2.00 | $6.00 | 500K | Jul 2026 |
| 29 | GLM-5.1Z.ai (GLM) | Active | 9 moAzure AI Foundry | 1464 | 89.9% | 1508 | – | – | – | Apr 2026 |
| 30 | Claude Sonnet 5Anthropic | Active | ≥ 2 moGoogle Vertex AI | 1461 | 90.5% | 1541 | $2.00 | $10.00 | 1M | Jun 2026 |
| 31 | Kimi K2.6Moonshot AI | Preview | ≥ 6 moAzure AI Foundry | 1461 | 90.8% | 1509 | – | – | 262.1K | – |
| 32 | DeepSeek-V4-ProDeepSeek | Active | 16 moAzure AI Foundry | 1458 | 91.7% | 1447 | $1.74 | $3.48 | 1M | – |
| 33 | GLM-5Z.ai (GLM) | Active | 7 daysGoogle Vertex AI | 1458 | 87.8% | 1434 | – | – | – | Feb 2026 |
| 34 | Claude Sonnet 4.5Anthropic | Active | 47 daysClaude API | 1456 | 82.3% | 1393 | $3.00 | $15.00 | 200K | Sep 2025 |
| 35 | Qwen3.7-PlusAlibaba Cloud (Qwen) | Active | None set | 1456 | 87.9% | – | $0.32 | $1.28 | – | – |
| 36 | GPT-6 SolOpenAI | Active | ≥ 11 moAmazon Bedrock | 1456 | 94.3% | 1688 | $2.00 | $10.00 | 1.1M | Sep 2026 |
| 37 | GPT-5.1OpenAI | Active | 6 moOpenAI API | 1456 | 87.6% | 1395 | $1.25 | $10.00 | 400K | Nov 2025 |
| 38 | Gemini 3.5 Flash-LiteGoogle | Active | ≥ 9 moGoogle Vertex AI | 1455 | 83.3% | 1440 | $0.30 | $2.50 | 1M | Jul 2026 |
| 39 | Grok 4.6xAI | Active | ≥ 10 moAmazon Bedrock | 1454 | 94.0% | 1617 | $2.00 | $6.00 | 500K | Aug 2026 |
| 40 | GPT-5.6 LunaOpenAI | Active | ≥ 9 moAmazon Bedrock | 1452 | 91.6% | – | $0.20 | $1.20 | 1.1M | Jul 2026 |
| 41 | Claude Opus 4.1Anthropic | Active | 3 moAmazon Bedrock | 1450 | 77.3% | 1390 | – | – | 200K | Aug 2025 |
| 42 | GPT-5.4 MiniOpenAI | Active | 11 moAzure AI Foundry | 1447 | 86.9% | 1397 | $0.75 | $4.50 | 400K | Mar 2026 |
| 43 | Gemini 2.5 ProGoogle | Legacy | 6 daysGoogle Vertex AI | 1446 | 85.3% | 1227 | $1.25 | $10.00 | 1M | Jun 2025 |
| 44 | Qwen3.6-PlusAlibaba Cloud (Qwen) | Active | None set | 1443 | 88.4% | 1461 | $0.50 | $3.00 | – | – |
| 45 | Grok 4.7xAI | Active | ≥ 11 moAmazon Bedrock | 1443 | 92.7% | 1639 | $2.00 | $6.00 | 500K | Sep 2026 |
| 46 | GPT-6 LunaOpenAI | Active | ≥ 11 moAmazon Bedrock | 1443 | 90.5% | 1581 | $0.10 | $0.50 | 1.1M | Sep 2026 |
| 47 | Qwen3.5-397B-A17BAlibaba Cloud (Qwen) | Active | 9 moAzure AI Foundry | 1442 | 86.4% | 1400 | – | – | 262.1K | – |
| 48 | Grok 4.3xAI | Active | ≥ 8 moAmazon Bedrock | 1442 | 88.8% | 1357 | $1.25 | $2.50 | 1M | – |
| 49 | GLM-4.7Z.ai (GLM) | Active | 7 daysGoogle Vertex AI | 1441 | 83.3% | 1435 | – | – | – | Dec 2025 |
| 50 | MiniMax-M3MiniMax | Active | None set | 1441 | 90.9% | 1482 | $0.30 | $1.20 | 1M | – |
| 51 | GPT-5.2OpenAI | Active | 8 moAzure AI Foundry | 1437 | 91.4% | 1416 | $1.75 | $14.00 | 400K | Dec 2025 |
| 52 | DeepSeek-V4-FlashDeepSeek | Active | 8 daysTogether AI | 1436 | 91.0% | 1582 | $0.44 | $1.32 | 1M | – |
| 53 | GPT-5OpenAI | Active | 58 daysOpenAI API | 1435 | 86.2% | 1418 | $1.25 | $10.00 | 400K | Aug 2025 |
| 54 | o3OpenAI | Active | 36 daysAzure AI Foundry | 1432 | 81.8% | – | $2.00 | $8.00 | 200K | Apr 2025 |
| 55 | Grok 4.1 Fast (Reasoning)xAI | Active | None set | 1430 | – | 1242 | – | – | 2M | Nov 2025 |
| 56 | Mistral Large 4Mistral AI | Preview | None set | 1429 | – | 1541 | $0.68 | $2.09 | 1M | Oct 2026 |
| 57 | Mistral Medium 3.5Mistral AI | Active | None set | 1426 | – | 1264 | $1.50 | $7.50 | 256K | Apr 2026 |
| 58 | Claude Opus 4Anthropic | Active | None set | 1426 | 76.3% | – | – | – | 200K | May 2025 |
| 59 | DeepSeek-V3.2DeepSeek | Active | 7 daysGoogle Vertex AI | 1425 | – | 1325 | $0.58 | $1.68 | – | Dec 2025 |
| 60 | Qwen3-235B-A22B-Instruct-2507Alibaba Cloud (Qwen) | Active | 7 daysGoogle Vertex AI | 1422 | – | – | $0.22 | $0.88 | 262.1K | Jul 2025 |
Arena (arena.ai), https://arena.ai/leaderboard. Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource]. Price is the provider’s own list price per 1M tokens on the model’s first-party platform; each model page lists every platform’s price. Blended price counts three input tokens to one output token. TeqEngine's measured agent work joins as a column when results publish.
Arena Text
Arena · rating · higher is better
GPQA Diamond
Epoch AI · percent · higher is better
Arena WebDev
Arena · rating · higher is better
Lowest price among the strongest 30
Blended $ per 1M tokens · lower is better
Largest context window
Tokens · current models
Arena (arena.ai), https://arena.ai/leaderboard. Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource]. Price is the provider’s own list price per 1M tokens on the model’s first-party platform; each model page lists every platform’s price. Blended price counts three input tokens to one output token. TeqEngine's measured agent work joins as a column when results publish.
Arena Text against blended price
Nothing scores higher for lessBeaten on both
Nothing scores higher for less
| Model | Arena Text | Blended |
|---|---|---|
| Gemini 4 Argon | 1525 | $4.00 |
| Gemini 3.8 Flash | 1497 | $1.50 |
| GLM-5.3-Flash | 1475 | $0.238 |
| GPT-6 Luna | 1443 | $0.20 |
| GLM-4.7-Flash | 1365 | $0.153 |
| Gemma 3 12B IT | 1342 | $0.14 |
| GPT-5 nano | 1337 | $0.138 |
| gpt-oss-20b | 1317 | $0.128 |
| Gemma 3 4B IT | 1303 | $0.05 |
Arena (arena.ai), https://arena.ai/leaderboard. Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource]. Price is the provider’s own list price per 1M tokens on the model’s first-party platform; each model page lists every platform’s price. Blended price counts three input tokens to one output token. TeqEngine's measured agent work joins as a column when results publish.