Skip to content
IFAI

Model rankings

866 models · 5 boards · 2 sources

WebDev arena · People compare web pages built by two models · Published November 8, 2026

Best web builder
Claude Opus 5.5 (Max)Anthropic1,818Score
Best open model
Kimi K3.1 Pro (Max)Moonshot AI1,687Score
Best from China
Kimi K3.1 Pro (Max)Moonshot AI1,687Score
Most voted
GPT-6 Luna (Max)OpenAI30,097Votes
Top three by Score
1Claude Opus 5.5 (Max)Anthropic1,818
Top three open models
1Kimi K3.1 Pro (Max)Moonshot AI · #9 overall1,687
2DeepSeek-V4.1-Flash-20261017 (Max)DeepSeek · #14 overall1,664
3Kimi K3 (Max)Moonshot AI · #15 overall1,657
This week's changes
Gemini 4 Carbon (High)Intelligence · #1New
GPT-6.1 Terra (Max)Intelligence · #6New
Fish Audio Drama 3Speech · #6New
P-Video-2-ProVideo · #11New
AssemblyAI-TTS-1Speech · #14New
RankChangeModelScore95% intervalVotesPrice $/M
1NewKimi K3.1 Pro (Max)OpenMoonshot AI1,687±104,651$4.0
2▼ 1DeepSeek-V4.1-Flash-20261017 (Max)OpenDeepSeek · 1 more variant1,664±713,449—
3▼ 1Kimi K3 (Max)OpenMoonshot AI1,657±622,147$6.0
4▼ 1MiMo-V2.6-FlashOpenXiaomi1,636±88,651$0.18
5▼ 1Hy4 previewOpenTencent1,632±104,599—
6—MiMo-V2.6-ProOpenXiaomi1,624±88,481$0.54
7▼ 2Qwen3.8-Flash-NextOpenAlibaba1,624±711,008—
8▼ 1GLM-5.3 (Max)OpenZ.ai1,622±712,326$2.2
9▼ 1GLM-5.3-FlashOpenZ.ai1,609±623,253$0.24
10▼ 1GLM-5.2 (Max)OpenZ.ai1,608±617,926—
11▼ 1Qwen3.8-27BOpenAlibaba1,591±620,482$1.1
12▼ 1MiniMax-M3.1-FlashOpenMiniMax1,582±104,516$0.26
13—DeepSeek-V4-Flash (High)OpenDeepSeek1,582±94,839—
14▼ 2DeepSeek-V4-Pro (High, 20260813)OpenDeepSeek · 1 more variant1,581±104,612—
15NewMistral Large 4OpenMistral1,563±161,466$1.0
16▼ 2Muse Spark Open (Max)OpenMeta1,528±151,799$2.0
17▼ 2Kimi K2.6OpenMoonshot AI1,509±710,012—
18▼ 2GLM-5.1OpenZ.ai1,508±710,540—
19▼ 2Hy3OpenTencent1,506±87,240—
20▼ 2MiniMax-M3OpenMiniMax1,484±521,059$0.52
21▼ 2MiMo-V2.5-ProOpenXiaomi1,479±522,092—
22▼ 2Kimi K2.7 CodeOpenMoonshot AI1,474±94,891—
23▼ 2DeepSeek-V4-Pro Preview (High)OpenDeepSeek1,463±612,952—
24▼ 2MiMo-V2.5OpenXiaomi1,438±614,634—
25▼ 1Beam 1.0 501B A23BOpenReflection1,438±142,522—
26▼ 3Kimi K2.5 (Thinking)OpenMoonshot AI · 1 more variant1,436±519,804—
27▼ 2GLM-5OpenZ.ai1,435±87,419—
28▼ 2GLM-4.7OpenZ.ai1,435±123,975—
29▼ 2DeepSeek-V4-Flash Preview (High)OpenDeepSeek1,431±79,983—
30▼ 2InklingOpenThinking Machines1,414±617,184$1.8
31▼ 2Inkling-SmallOpenThinking Machines1,407±95,227$0.52
32▼ 2Qwen3.5-397B-A17BOpenAlibaba1,400±520,859—
33▼ 2NVIDIA-Nemotron-3.5-Super-VL-120B-A12B (NVFP4)OpenNVIDIA1,399±114,732—
34▼ 2MiniMax-M2.7OpenMiniMax1,398±615,742—
35▼ 2MiniMax-M2.5OpenMiniMax1,386±87,576—
36▼ 2MiniMax-M2.1 PreviewOpenMiniMax1,384±107,579—
37▼ 2Gemma 4 31BOpenGoogle1,369±616,301$0.000
38▼ 2DeepSeek-V3.2 (Thinking)OpenDeepSeek · 1 more variant1,363±96,856—
39▼ 2Qwen3.5-122B-A10BOpenAlibaba1,360±88,188—
40▼ 2Gemma 4 26B A4BOpenGoogle1,359±171,491—
41▼ 1Hy3 previewOpenTencent1,359±171,424—
42▼ 3Qwen3.5-27BOpenAlibaba1,357±87,738—
43▼ 2Muse GlimmerOpenMeta1,355±161,634$0.52
44▼ 2Laguna M.1OpenPoolside1,349±95,315—
45▼ 2GLM-4.6OpenZ.ai1,342±116,680—
46▼ 2MiMo-V2-Flash (Non-Thinking)OpenXiaomi · 1 more variant1,331±105,647—
47▼ 2Kimi K2 Thinking TurboOpenMoonshot AI1,322±712,959—
48▼ 2Laguna XS.2OpenPoolside1,303±114,166—
49▼ 2MiniMax-M2OpenMiniMax1,299±116,602—
50▼ 2Qwen3-Coder-480B-A35B-InstructOpenAlibaba1,275±812,576—

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 “—”.