Skip to content
IFAI

Model rankings

1037 models · 5 boards · 2 sources

Artificial AnalysisIntelligence IndexArenaText arena

Intelligence Index · A composite of several evaluations · Updated October 19, 2026

Smartest
Claude Opus 5.5 (Max, Default Fallback)Anthropic57.6Intelligence
Best open model
MiMo-V2.6-ProXiaomi46.3Intelligence
Best value
Agnes 2.5 Pro BetaSapiens AI235 pts/$Value
Best from China
MiMo-V2.6-ProXiaomi46.3Intelligence
Top three open models
1MiMo-V2.6-ProXiaomi · #11 overall46.3
2GLM-5.3 (Max)Z AI · #13 overall44.8
3Step 5 PreviewStepFun · #14 overall43.7
This week's changes
Muse Wonder Full (Max)Intelligence · #3New
FLUX 3 Max (720p)Video · #4New
Muse Wonder Full (Max)Coding · #6New
Muse Wonder (Max)Intelligence · #7New
Kimi K3.1 Pro (Max)Coding · #7New
RankChangeModelIntelligencePrice $/MSpeed tok/sFirst token
1NewMuse Spark Open (Xhigh)OpenMeta38.9$2.01410.8 s
2NewNemotron 3.5 Super VLOpenNVIDIA26.5$0.314220.8 s
3▼ 2Inkling SmallOpenThinking Machines25.7$0.521992.1 s
4▼ 2Inkling (Xhigh)OpenThinking Machines25.0$1.82161.8 s
5▼ 2Nemotron 3 Ultra 550B A55B (Reasoning)OpenNVIDIA22.9$1.11441.2 s
6NewNemotron 3.5 Bolt 74B-A9BOpenNVIDIA21.4———
7▼ 3Muse Glimmer (High)OpenMeta17.5$0.521430.9 s
8▼ 3Gemma 4 26B A4B (Reasoning)OpenGoogle · 1 more variant16.7$0.17711.0 s
9▼ 3Gemma 4 31B (Reasoning)OpenGoogle · 1 more variant14.7$0.000361.1 s
10▼ 3Gemma 4 12B (Reasoning)OpenGoogle · 1 more variant14.2$0.151192.3 s
11▼ 2Apriel-v1.6-15B-ThinkerOpenServiceNow13.4———
12▼ 2Nemotron 3.5 LightningOpenNVIDIA12.9$0.0952980.5 s
13▼ 1Granite 4.2 30BOpenIBM12.7$0.28770.9 s
14▼ 1Nemotron Cascade 2 30B A3BOpenNVIDIA11.7———
15▼ 1gpt-oss-120b (High)OpenOpenAI · 1 more variant11.6$0.242580.8 s
16▼ 1Granite 4.2 8BOpenIBM11.1$0.111200.8 s
17▼ 1Nemotron 3 Nano Omni 30B A3B ReasoningOpenNVIDIA10.3$0.452360.8 s
18—gpt-oss-20b (Low)OpenOpenAI · 1 more variant10.0$0.123020.7 s
19—DiffusionGemma 26B A4BOpenGoogle9.5———
20▼ 3Llama 4 MaverickOpenMeta9.3$0.42471.1 s
21▼ 1Granite 4.2 3BOpenIBM9.1$0.0532180.5 s
22▲ 1Gemma 4 E4B (Reasoning)OpenGoogle · 1 more variant8.9$0.040370.9 s
23▲ 1NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)OpenNVIDIA · 1 more variant8.9$0.0882530.9 s
24▲ 1LFM2.5-2.6BOpenLiquid AI8.4———
25▲ 2Gemma 4 E2B (Reasoning)OpenGoogle · 1 more variant7.8———
26▲ 2Llama 3.3 Instruct 70BOpenMeta7.7$0.71931.7 s
27▲ 2Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)OpenNVIDIA7.5———
28▲ 2NVIDIA Nemotron Nano 12B v2 VL (Reasoning)OpenNVIDIA · 1 more variant7.5———
29▲ 3NVIDIA Nemotron Nano 9B V2 (Reasoning)OpenNVIDIA · 1 more variant7.4$0.0701950.6 s
30▲ 4Llama 3.1 Instruct 405BOpenMeta7.3———
31▲ 4LFM2.5-8B-A1BOpenLiquid AI7.2———
32▼ 6Llama 4 ScoutOpenMeta6.5$0.301170.8 s
33▲ 7R1 1776OpenPerplexity6.4———
34▲ 7Llama 3.2 Instruct 90B (Vision)OpenMeta6.4———
35▲ 7Phi-4 Mini InstructOpenMicrosoft6.3$0.13440.8 s
36▲ 7Granite 4.0 H SmallOpenIBM6.0———
37▲ 7LFM2 24B A2BOpenLiquid AI5.9———
38▲ 7Phi-4OpenMicrosoft5.9$0.22412.6 s
39▲ 10Phi-4 Multimodal InstructOpenMicrosoft5.8$0.14——
40▲ 14Llama 3.2 Instruct 11B (Vision)OpenMeta5.4$0.35191.3 s
41▲ 17LFM2.5-1.2B-ThinkingOpenLiquid AI5.2———
42▲ 18LFM2.5-1.2B-InstructOpenLiquid AI5.2———
43▲ 18Granite 4.0 H 1BOpenIBM5.2———
44▲ 18Gemma 3 270MOpenGoogle5.1———
45▲ 18Granite 4.0 MicroOpenIBM5.1———
46▲ 23Granite 4.0 350MOpenIBM4.8———
46▲ 23Granite 4.0 H 350MOpenIBM4.8———
46▲ 23LFM2.5-VL-1.6BOpenLiquid AI4.8———
49▲ 30NVIDIA Nemotron 3 Nano 4BOpenNVIDIA4.8———

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 “—”.