Skip to content
IFAI

Model rankings

866 models · 5 boards · 2 sources

Artificial AnalysisIntelligence IndexArenaText arena

Intelligence Index · A composite of several evaluations · Updated December 9, 2025

Smartest
Gemini 3 Pro Preview (high)Google72.9Intelligence
Best open model
Kimi K2 ThinkingKimi67.0Intelligence
Best value
Gemma 3n E4B InstructGoogle620 pts/$Value
Best from China
Kimi K2 ThinkingKimi67.0Intelligence
Top three by Intelligence
3GPT-5.1 (high)OpenAI69.7
Top three open models
1Kimi K2 ThinkingKimi · #6 overall67.0
2DeepSeek V3.2 (Reasoning)DeepSeek · #8 overall65.9
3MiniMax-M2MiniMax · #18 overall61.4
This week's changes
gpt-5.2Coding · #5New
GPT-5.1 Codex (high)Intelligence · #7New
DeepSeek V3.2 (Reasoning)Intelligence · #8New
Z-Image TurboImage · #9New
Nova 2.0 Pro Preview (medium)Intelligence · #16New
RankChangeModelIntelligencePrice $/MSpeed tok/sFirst token
1—Kimi K2 ThinkingOpenKimi67.0$1.1——
2NewDeepSeek V3.2 (Reasoning)OpenDeepSeek65.9$0.32——
3▼ 1MiniMax-M2OpenMiniMax61.4$0.52——
4NewDeepSeek V3.2 SpecialeOpenDeepSeek58.6$0.32——
5▼ 2DeepSeek V3.1 Terminus (Reasoning)OpenDeepSeek57.7$0.80——
6▼ 2Qwen3 235B A22B 2507 (Reasoning)OpenAlibaba57.5$2.6——
7▼ 2Doubao Seed CodeClosedByteDance Seed57.2$0.41——
8▼ 2DeepSeek V3.2 Exp (Reasoning)OpenDeepSeek56.9$0.32——
9▼ 2GLM-4.6 (Reasoning)OpenZ AI56.0$1.0——
10▼ 2Qwen3 Max ThinkingClosedAlibaba55.8$2.4——
11▼ 2Qwen3 MaxClosedAlibaba55.2$2.4——
12▼ 2Qwen3 VL 235B A22B (Reasoning)OpenAlibaba54.4$2.6——
13▼ 2Qwen3 Next 80B A3B (Reasoning)OpenAlibaba54.3$1.9——
14▼ 2DeepSeek V3.1 (Reasoning)OpenDeepSeek54.0$0.65——
15NewERNIE 5.0 Thinking PreviewClosedBaidu53.2$1.5——
16NewDeepSeek V3.2 (Non-reasoning)OpenDeepSeek52.4$0.32——
17▼ 4DeepSeek R1 0528 (May '25)OpenDeepSeek52.0$1.7——
18▼ 4Qwen3 VL 32B (Reasoning)OpenAlibaba51.9$2.6——
19▼ 4Seed-OSS-36B-InstructOpenByteDance Seed51.6$0.30——
20▼ 4GLM-4.5 (Reasoning)OpenZ AI51.3$1.0——
21▼ 4Kimi K2 0905OpenKimi50.4$1.4——
22▼ 4GLM-4.5-AirOpenZ AI48.8$0.43——
23▼ 4Qwen3 Max (Preview)ClosedAlibaba48.6$2.4——
24▼ 4Kimi K2OpenKimi48.2$1.1——
25▼ 4Qwen3 30B A3B 2507 (Reasoning)OpenAlibaba46.4$0.75——
26▼ 4DeepSeek V3.2 Exp (Non-reasoning)OpenDeepSeek46.3$0.32——
27▼ 4MiniMax M1 80kOpenMiniMax46.2$0.83——
28▼ 4DeepSeek V3.1 Terminus (Non-reasoning)OpenDeepSeek45.7$0.72——
29▼ 4Qwen3 235B A22B 2507 InstructOpenAlibaba45.3$1.2——
30▼ 4Qwen3 VL 30B A3B (Reasoning)OpenAlibaba45.3$0.75——
31▼ 4Qwen3 Next 80B A3B InstructOpenAlibaba44.8$0.88——
32▼ 4Ling-1TOpenInclusionAI44.8$1.00——
33▼ 4DeepSeek V3.1 (Non-reasoning)OpenDeepSeek44.8$0.84——
34▼ 4GLM-4.6 (Non-reasoning)OpenZ AI44.7$1.0——
35▼ 4Qwen3 VL 235B A22B InstructOpenAlibaba44.2$1.2——
36▼ 4DeepSeek R1 (Jan '25)OpenDeepSeek43.8$2.0——
37▼ 4Qwen3 4B 2507 (Reasoning)OpenAlibaba43.4$0.000——
38▼ 4Qwen3 Coder 480B A35B InstructOpenAlibaba42.3$3.0——
39▼ 4Ring-1TOpenInclusionAI41.8$0.99——
40▼ 4Qwen3 235B A22B (Reasoning)OpenAlibaba41.7$2.6——
41▼ 4DeepSeek V3 0324OpenDeepSeek41.3$1.2——
42▼ 4Qwen3 VL 32B InstructOpenAlibaba41.0$1.2——
43▼ 4MiniMax M1 40kOpenMiniMax40.0$0.83——
44▼ 4Qwen3 Omni 30B A3B (Reasoning)OpenAlibaba40.0$0.43——
45▼ 4Ring-flash-2.0OpenInclusionAI39.5$0.25——
46▼ 4Qwen3 32B (Reasoning)OpenAlibaba38.7$2.6——
47▼ 4Qwen3 VL 30B A3B InstructOpenAlibaba38.5$0.35——
48▼ 4Ling-flash-2.0OpenInclusionAI38.3$0.25——
49▼ 4QwQ 32BOpenAlibaba37.9$0.47——
50▼ 4GLM-4.5V (Reasoning)OpenZ AI37.0$0.85——

Intelligence over time

Each maker's best model at the time on the Artificial Analysis Intelligence Index, by the month it was released.

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

Arena data comes from the leaderboard dataset it publishes on Hugging Face (CC BY 4.0). Only each board's overall leaderboard is used; scores and ranks are unchanged. dataset · CC BY 4.0

Artificial Analysis data comes from its API (artificialanalysis.ai). Scores are unchanged; ranks follow the scores. artificialanalysis.ai

Open or closed follows the license Arena gives; prices and Artificial Analysis intelligence show only where model names match, otherwise “—”.