BenchLeader
Menu
Best LLMsChoose a modelCoding plansLeaderboardsModelsBenchmarksProvidersReleases12

Design Arena Website leaderboard

As of 9 Nov 2026, the top score on Design Arena Website is 1368 by Muse Wonder (Meta), ahead of Muse Spark 1.3 at 1353, across 189 model configurations with a published result (Design Arena, result dated 9 Nov 2026). Design Arena asks two models to build a website from the same prompt, renders both side by side, and lets the visitor pick the better one. Design Arena Website is shown for reference only and does not count toward the BenchLeader Index.

Key facts (as of 9 Nov 2026)

  • Muse Wonder holds the top Design Arena Website score at 1368 (Design Arena, 9 Nov 2026).
  • Muse Spark 1.3 is second on Design Arena Website at 1353 (Design Arena, 9 Nov 2026).
  • Design Arena Website: Two models build a website from the same prompt, the visitor votes for the better one, and the votes become a Bradley-Terry rating. Models with fewer than 200 battles are left out (Design Arena).
  • 189 model configurations have a published Design Arena Website result on BenchLeader (BenchLeader, 9 Nov 2026).
  • Design Arena Website does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).

Two models build a website from the same prompt, the visitor votes for the better one, and the votes become a Bradley-Terry rating. Models with fewer than 200 battles are left out.

Published by
Design Arena
Category
Coding
Index weight
Reference only
Models
189
Data as of
9 Nov 2026

© Arcada Labs. Design Arena (https://www.designarena.ai/leaderboard). Ratings read daily from the public leaderboard and shown with credit and a link; not endorsed by Arcada Labs.

What the test looks like

Design Arena asks two models to build a website from the same prompt, renders both side by side, and lets the visitor pick the better one.

How it is scored

Bradley-Terry rating from the pairwise votes (Design Arena calls it Elo). Rows need at least 200 battles here.

What to keep in mind

Votes reward how a first draft looks, not whether the code is maintainable. Overlaps in spirit with LMArena WebDev, with a different crowd and model list. Kept out of the index.

Top 15
  1. 1Muse Wonder (max)1368
  2. 2Muse Spark 1.3 (xhigh)1353
  3. 3Muse Spark 1.3 (max)1343
  4. 4Inkling 2 Small1334
  5. 5GPT-6 Astra (max)1333
  6. 6Kimi K31332
  7. 7GPT-6 Astra (xhigh)1331
  8. 8Muse Ember1330
  9. 9Claude Opus 5.51325
  10. 10Kimi K3.1 Pro1320
  11. 11GPT-5.6 Sol (xhigh)1317
  12. 12Gemini 4 Argon1315
  13. 13Claude Sonnet 5.51315
  14. 14GPT-6.1 Sol1310
  15. 15Muse Spark 1.21310
189 of 189
#
1Muse Wondermax136861.92026-11-09
2Muse Spark 1.3xhigh135358.12026-11-09
3Muse Spark 1.3max134361.72026-11-09
4NewInkling 2 Small133458.32026-11-09
5GPT-6 Astramax133367.42026-11-09
6Kimi K3133259.52026-11-09
7GPT-6 Astraxhigh133166.22026-11-09
8NewMuse Ember1330–2026-11-09
9Claude Opus 5.5132563.22026-11-09
10Kimi K3.1 Pro132061.22026-11-09
11GPT-5.6 Solxhigh131764.22026-11-09
12Gemini 4 Argon131563.32026-11-09
13Claude Sonnet 5.51315–2026-11-09
14GPT-6.1 Sol131062.22026-11-09
15Muse Spark 1.21310–2026-11-09
16Claude Fable 5.1130666.02026-11-09
17Claude Opus 5130665.22026-11-09
18GPT-5.6 Solmedium130561.42026-11-09
19MiMo-V2.6-Pro130557.12026-11-09
20DeepSeek V4.1 Flash1305–2026-11-09
21Muse Spark Openxhigh130253.62026-11-09
22Gemini 3.7 Flash1300–2026-11-09
23Gemini 3.8 Flash129754.32026-11-09
24Gemini 3.6 Flash129549.62026-11-09
25DeepSeek V4.1 Flash 10171295–2026-11-09
26Claude Fable 5129266.62026-11-09
27GPT-6.1 Terra1288–2026-11-09
28Claude Opus 4.6128762.32026-11-09
29Ling 3.1 Flash128650.82026-11-09
30GLM 5.31286–2026-11-09
31GPT-6 Sol1285–2026-11-09
32Grok 4.6128456.82026-11-09
33Claude Opus 4.6thinking128362.62026-11-09
34Claude Sonnet 4.6128158.22026-11-09
35GLM 5.2128151.52026-11-09
36Grok 4.5127759.92026-11-09
37MiniMax M3.1 Flash127753.92026-11-09
38Qwen3.7max127260.32026-11-09
39Qwen3.8 Maxmax127258.92026-11-09
40Claude Sonnet 5127056.02026-11-09
41Muse Spark 1.1126762.32026-11-09
42GLM 5.3 Flash126756.82026-11-09
43Gemini 3.5 Flash126648.82026-11-09
44Qwen3.7 Plus126557.72026-11-09
45Kimi K2.6126555.22026-11-09
46GPT-6 Luna1265–2026-11-09
47MiMo-V2.5-Pro126455.62026-11-09
48GPT-5.6 Terra126451.72026-11-09
49MiMo-V2.5126356.32026-11-09
50GLM 5.1126355.82026-11-09
51Kimi K2.7 Code126054.22026-11-09
52Muse Spark125864.22026-11-09
53Claude Haiku 5.51257–2026-11-09
54GLM 5 Turbo1257–2026-11-09
55Claude Opus 4.8125660.12026-11-09
56Hy4125655.32026-11-09
57Mistral Large 4125452.22026-11-09
58GPT-5.5125361.72026-11-09
59MiniMax M3125356.32026-11-09
60GPT-5.6 Luna125247.72026-11-09

More coding benchmarks · All LLM benchmarks

Related benchmarks

Top head-to-heads on Design Arena Website

Cite as: BenchLeader, “Design Arena Website leaderboard”, data as of 9 Nov 2026, https://www.benchleader.com/benchmarks/designarena_website.

Design Arena Website: questions

What does Design Arena Website measure?
Design Arena asks two models to build a website from the same prompt, renders both side by side, and lets the visitor pick the better one. Scores are reported in Elo-style ratings from pairwise comparisons; higher is better.
What is the top score on Design Arena Website?
As of 9 Nov 2026, the highest score on Design Arena Website is 1368 by Muse Wonder (result dated 9 Nov 2026), ahead of Muse Spark 1.3 at 1353. Scores come from Design Arena and are re-read daily.
Which AI model leads Design Arena Website?
Muse Wonder leads Design Arena Website with 1368 as of 9 Nov 2026 (result dated 9 Nov 2026), ahead of Muse Spark 1.3 at 1353.
How many models have Design Arena Website results?
189 model configurations have a Design Arena Website result on BenchLeader, all taken from Design Arena.
Who runs Design Arena Website and how often is it updated?
Design Arena Website is published by Design Arena. BenchLeader re-reads the published results once a day and records the date each result was published.
Does Design Arena Website count toward the BenchLeader Index?
No. Design Arena Website is shown for reference but left out of the composite index.