Model rankings
277 models · 2 boards · 2 sources
Smartest
Best open model
Best from China
Top three open models
DeepSeek R1DeepSeek · #3 overall60.2
DeepSeek R1 Distill Qwen 32BDeepSeek · #6 overall51.9
DeepSeek R1 Distill Qwen 14BDeepSeek · #8 overall49.0
This week's changes
SonarIntelligence · #15New
Sonar ProIntelligence · #16New
Nova LiteIntelligence · #42New
Jamba 1.6 LargeIntelligence · #45New
Phi-4 Multimodal InstructIntelligence · #62New
| Rank | Change | Model | Intelligence | Price $/M | Speed tok/s | First token |
|---|---|---|---|---|---|---|
| 1 | — | o3-mini (high)ClosedOpenAI · 1 more variant | 65.5 | — | — | — |
| 2 | — | o1ClosedOpenAI | 61.9 | — | — | — |
| 3 | — | DeepSeek R1OpenDeepSeek | 60.2 | — | — | — |
| 4 | — | Claude 3.7 Sonnet (Extended Thinking)ClosedAnthropic · 1 more variant | 57.4 | — | — | — |
| 5 | — | o1-miniClosedOpenAI | 53.8 | — | — | — |
| 6 | — | DeepSeek R1 Distill Qwen 32BOpenDeepSeek | 51.9 | — | — | — |
| 7 | — | Gemini 2.0 Pro Experimental (Feb '25)ClosedGoogle | 49.2 | — | — | — |
| 8 | — | DeepSeek R1 Distill Qwen 14BOpenDeepSeek | 49.0 | — | — | — |
| 9 | — | DeepSeek R1 Distill Llama 70BOpenDeepSeek | 48.2 | — | — | — |
| 10 | — | Gemini 2.0 Flash (Feb '25)ClosedGoogle | 48.1 | — | — | — |
| 11 | — | DeepSeek V3OpenDeepSeek | 45.7 | — | — | — |
| 12 | — | Qwen2.5 MaxClosedAlibaba | 45.3 | — | — | — |
| 13 | — | Gemini 1.5 Pro (Sep '24)ClosedGoogle · 1 more variant | 44.6 | — | — | — |
| 14 | — | Claude 3.5 Sonnet (Oct '24)ClosedAnthropic | 44.3 | — | — | — |
| 15 | New | SonarClosedPerplexity | 44.1 | — | — | — |
| 16 | New | Sonar ProClosedPerplexity | 42.8 | — | — | — |
| 17 | ▼ 2 | QwQ 32B-PreviewOpenAlibaba | 42.6 | — | — | — |
| 18 | ▼ 2 | Gemini 2.0 Flash-Lite (Preview)ClosedGoogle | 41.9 | — | — | — |
| 19 | ▼ 2 | GPT-4o (Nov '24)ClosedOpenAI · 3 more variants | 41.5 | — | — | — |
| 20 | ▼ 2 | Llama 3.3 Instruct 70BOpenMeta | 41.1 | — | — | — |
| 21 | ▼ 2 | Llama 3.1 Instruct 405BOpenMeta | 40.5 | — | — | — |
| 22 | ▼ 2 | Qwen2.5 Instruct 72BOpenAlibaba | 40.3 | — | — | — |
| 23 | ▼ 2 | Phi-4OpenMicrosoft Azure | 40.2 | — | — | — |
| 24 | ▼ 2 | Llama 3.1 Tulu3 405BOpenAllen Institute for AI | 39.9 | — | — | — |
| 25 | ▼ 2 | MiniMax-Text-01OpenMiniMax | 39.9 | — | — | — |
| 26 | ▼ 2 | Mistral Large 2 (Nov '24)OpenMistral · 1 more variant | 38.3 | — | — | — |
| 27 | ▼ 2 | Grok BetaClosedxAI | 37.5 | — | — | — |
| 28 | ▼ 2 | Pixtral LargeOpenMistral | 37.4 | — | — | — |
| 29 | ▼ 2 | Qwen2.5 Instruct 32BOpenAlibaba | 37.4 | — | — | — |
| 30 | ▼ 2 | Llama 3.1 Nemotron Instruct 70BOpenNVIDIA | 37.3 | — | — | — |
| 31 | ▼ 2 | Nova ProClosedAmazon | 37.1 | — | — | — |
| 32 | ▼ 2 | Qwen2.5 Coder Instruct 32BOpenAlibaba | 36.3 | — | — | — |
| 33 | ▼ 2 | GPT-4o miniClosedOpenAI | 35.7 | — | — | — |
| 34 | ▼ 2 | Llama 3.1 Instruct 70BOpenMeta | 35.4 | — | — | — |
| 35 | ▼ 2 | Mistral Small 3OpenMistral | 35.3 | — | — | — |
| 36 | ▼ 2 | Claude 3 OpusClosedAnthropic | 35.1 | — | — | — |
| 37 | ▼ 2 | Claude 3.5 HaikuClosedAnthropic | 34.6 | — | — | — |
| 38 | ▼ 2 | DeepSeek R1 Distill Llama 8BOpenDeepSeek | 34.0 | — | — | — |
| 39 | ▼ 2 | Qwen TurboClosedAlibaba | 33.6 | — | — | — |
| 40 | ▼ 2 | Llama 3.2 Instruct 90B (Vision)OpenMeta | 33.4 | — | — | — |
| 41 | ▼ 2 | Qwen2 Instruct 72BOpenAlibaba | 32.6 | — | — | — |
| 42 | New | Nova LiteClosedAmazon | 32.5 | — | — | — |
| 43 | ▼ 3 | Mistral SabaClosedMistral | 32.2 | — | — | — |
| 44 | ▼ 3 | Jamba 1.5 LargeOpenAI21 Labs | 29.3 | — | — | — |
| 45 | New | Jamba 1.6 LargeOpenAI21 Labs | 28.8 | — | — | — |
| 46 | ▼ 4 | Gemini 1.5 Flash (May '24)ClosedGoogle | 28.4 | — | — | — |
| 47 | ▼ 4 | Nova MicroClosedAmazon | 28.3 | — | — | — |
| 48 | ▼ 4 | Yi-LargeClosed01.AI | 27.9 | — | — | — |
| 49 | ▼ 4 | Claude 3 SonnetClosedAnthropic | 27.8 | — | — | — |
| 50 | ▼ 4 | Codestral (Jan '25)ClosedMistral · 1 more variant | 27.6 | — | — | — |
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