Model rankings
215 models · 2 boards · 2 sources
Week of February 17, 2025
Smartest
Best open model
Best from China
Top three open models
DeepSeek R1DeepSeek · #3 overall60.2
DeepSeek V3DeepSeek · #6 overall45.7
QwQ 32B-PreviewAlibaba · #11 overall42.6
| Rank | Change | Model | Intelligence | Price $/M | Speed tok/s | First token |
|---|---|---|---|---|---|---|
| 1 | — | o3-miniClosedOpenAI | 62.9 | — | — | — |
| 2 | — | o1ClosedOpenAI | 61.9 | — | — | — |
| 3 | — | o1-miniClosedOpenAI | 53.8 | — | — | — |
| 4 | — | Gemini 2.0 Pro Experimental (Feb '25)ClosedGoogle | 46.5 | — | — | — |
| 5 | — | Gemini 1.5 Pro (Sep '24)ClosedGoogle | 44.6 | — | — | — |
| 6 | — | Claude 3.5 Sonnet (Oct '24)ClosedAnthropic | 44.3 | — | — | — |
| 7 | — | Gemini 2.0 Flash (Feb '25)ClosedGoogle | 44.2 | — | — | — |
| 8 | — | Gemini 2.0 Flash-Lite (Preview)ClosedGoogle | 41.9 | — | — | — |
| 9 | — | GPT-4o (Nov '24)ClosedOpenAI · 2 more variants | 41.5 | — | — | — |
| 10 | — | Llama 3.3 Instruct 70BOpenMeta | 41.1 | — | — | — |
| 11 | — | Llama 3.1 Instruct 405BOpenMeta | 40.5 | — | — | — |
| 12 | — | Llama 3.1 Nemotron Instruct 70BOpenNVIDIA | 37.3 | — | — | — |
| 13 | — | Nova ProClosedAmazon | 37.1 | — | — | — |
| 14 | — | GPT-4o miniClosedOpenAI | 35.7 | — | — | — |
| 15 | — | Llama 3.1 Instruct 70BOpenMeta | 35.4 | — | — | — |
| 16 | — | Claude 3 OpusClosedAnthropic | 35.1 | — | — | — |
| 17 | — | Claude 3.5 HaikuClosedAnthropic | 34.6 | — | — | — |
| 18 | — | Nova MicroClosedAmazon | 28.3 | — | — | — |
| 19 | — | Llama 3 Instruct 70BOpenMeta | 27.5 | — | — | — |
| 20 | — | Llama 3.1 Instruct 8BOpenMeta | 23.7 | — | — | — |
| 21 | — | Llama 3.2 Instruct 11B (Vision)OpenMeta | 22.2 | — | — | — |
Intelligence over time
Gained most · Anthropic +57.2Highest now · Anthropic 64.6Gap between highest and lowest 18.1 ptsTime range Nov 2025 – Nov 2026
64.6Anthropic60.0Google56.9OpenAI50.1Meta46.5Kimi