Skip to content
IFAI

Model rankings

1512 models · 5 boards · 2 sources

WebDev arena · People compare web pages built by two models · Published September 25, 2026

Best web builder
claude-opus-5.5-maxAnthropic1,827Score
Best open model
kimi-k3-maxMoonshot AI1,660Score
Best from China
qwen3.8-maxAlibaba1,672Score
Most voted
claude-haiku-4-5-20251001Anthropic29,211Votes
Top three by Score
1claude-opus-5.5-maxAnthropic1,827
2gpt-6-astra-maxOpenAI1,792
3claude-fable-5.1-maxAnthropic1,751
Top three open models
1kimi-k3-maxMoonshot AI · #8 overall1,660
2qwen3.8-flash-nextAlibaba · #10 overall1,636
3hy4-previewTencent · #11 overall1,631
This week's changes
claude-opus-5.5-maxCoding · #1New
Gemini 3.8 Flash TTSSpeech · #2New
Grok Imagine Image 2.0Image · #4New
gpt-6-sol-maxCoding · #5New
RankChangeModelScore95% intervalVotesPrice $/M
1Newclaude-opus-5.5-maxClosedAnthropic1,827±181,607—
2▼ 1gpt-6-astra-maxClosedOpenAI1,792±114,908—
3▼ 1claude-fable-5.1-maxClosedAnthropic1,751±105,313—
4▼ 1claude-opus-5-maxClosedAnthropic1,693±715,627—
5Newgpt-6-sol-maxClosedOpenAI1,681±152,019—
6▼ 2qwen3.8-maxClosedAlibaba · 1 more variant1,672±123,469$3.0
7▼ 1claude-opus-5-highClosedAnthropic1,662±618,930—
8▼ 3kimi-k3-maxOpenMoonshot AI1,660±714,739—
9▼ 2muse-spark-1.3-maxClosedMeta1,656±95,945—
10▼ 2qwen3.8-flash-nextOpenAlibaba1,636±104,980—
11▼ 1hy4-previewOpenTencent1,631±104,954—
12Newgrok-4.7-xhighClosedxAI1,629±142,047—
13Newclaude-fable-5-highClosedAnthropic1,627±713,086—
14▼ 3muse-spark-1.3 (xHigh)ClosedMeta1,626±95,043$2.0
15—deepseek-v4.1-flash-maxOpenDeepSeek1,621±113,815—
16▼ 4grok-4.6-highClosedxAI1,621±87,040—
17▼ 3glm-5.3-maxOpenZ.ai1,619±96,410—
18Newmimo-v2.6-proOpenXiaomi1,618±132,591$0.54
19▼ 6gpt-5.6-sol-xhigh (codex-harness)ClosedOpenAI1,617±615,764—
20▼ 4glm-5.3-flashOpenZ.ai1,613±96,665$0.24
21▼ 3glm-5.2-maxOpenZ.ai1,602±713,601—
22▼ 3gemini-3.7-flash-highClosedGoogle1,593±88,086—
23Newgpt-6-luna-maxClosedOpenAI1,593±141,903—
24▼ 7qwen3.8-27bOpenAlibaba1,591±710,236$1.1
25▼ 5deepseek-v4-pro-high-20260813OpenDeepSeek · 1 more variant1,582±94,897—
26▼ 5deepseek-v4-flash-highOpenDeepSeek · 1 more variant1,581±95,067—
27▼ 5gemini-3.8-flash-highClosedGoogle1,580±87,338—
28Newstep-5-preview-highClosedStepFun1,565±171,247—
29▼ 4claude-opus-4-7-highClosedAnthropic1,557±617,538—
30▼ 6claude-opus-4-7ClosedAnthropic1,556±616,945—
31▼ 8claude-opus-4-8-highClosedAnthropic1,555±617,611—
32▼ 6grok-4.5ClosedxAI1,552±711,302—
33▼ 6claude-opus-4-6-highClosedAnthropic1,546±519,527—
34▼ 6muse-spark-1.1ClosedMeta1,541±88,102—
35▼ 3claude-sonnet-5-highClosedAnthropic1,539±612,862—
36▼ 6claude-opus-4-6ClosedAnthropic1,537±520,788—
37▼ 6gemini-3.6-flash-highClosedGoogle1,536±78,857—
38▼ 9claude-opus-4-8ClosedAnthropic1,535±616,576—
39▼ 6muse-spark-1.2 (xHigh)ClosedMeta1,531±142,208—
40▼ 5claude-sonnet-4-6ClosedAnthropic1,521±522,980—
41▼ 7gpt-5.6-terra-xhigh (codex-harness)ClosedOpenAI1,521±711,194—
42▼ 6gpt-5.6-luna-xhigh (codex-harness)ClosedOpenAI1,520±711,419—
43▼ 6seed-2.1-pro-previewClosedByteDance1,518±612,270—
44▼ 6qwen3.7-max-20260517ClosedAlibaba1,515±78,931—
45▼ 5gpt-5.5-xhigh (codex-harness)ClosedOpenAI1,512±615,882—
46▼ 7hy3OpenTencent1,509±95,302—
47▼ 6kimi-k2.6OpenMoonshot AI1,509±710,202—
48▼ 6glm-5.1OpenZ.ai1,508±710,822—
49▼ 6gemini-3.5-flash-highClosedGoogle1,499±79,530—
50▼ 6claude-opus-4-5-20251101-high-32kClosedAnthropic1,493±811,035—

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 “—”.