Benchmark Almanac

DesignArena Image to HTML

GPT-6 Astra 1,304 leads since September 2026
Not in model cards

Models recreate a reference image or mockup as a working single-file HTML website. Users compare the rendered pages and vote for the closer, better one.

One of 40 DesignArena boards. DesignArena ranks entries with Bradley-Terry ratings from user votes.

Kind
Web and app front ends
Level
Everyday
Scored
Ranked by votes
Answers
Open-ended
Tasks
All public
Strongest model tested
Claude Fable 5.2 #1 on the capability index
Last model added
November 2026
Scores refreshed
November 12th, 2026
53
Entries on its own leaderboard, 51 models we track

GPT-6 Astra leads with 1,304

Best score per entry from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

53 entries. Scroll sideways for the rest. Bars start at 1,000.

  1. 1,304 GPT-6 Astra
  2. 1,273 Muse Ember
  3. 1,271 Gemini 4 Argon
  4. 1,269 Kimi K3
  5. 1,267 Claude Opus 5.5
  6. 1,264 Muse Wonder
  7. 1,257 Muse Spark 1.3
  8. 1,254 Kimi K3.1 Pro
  9. 1,247 Claude Fable 5.2
  10. 1,233 GPT-6 Sol
  11. 1,232 Claude Sonnet 5.5
  12. 1,231 GPT-5.6 Sol
  13. 1,231 GPT-6.1 Sol
  14. 1,230 Gemini 3.8 Flash
  15. 1,227 Gemini 3.7 Flash
  16. 1,227 GPT-6.1 Terra
  17. 1,226 DeepSeek V4.1 Flash
  18. 1,218 Muse Spark 1.2
  19. 1,218 Grok 4.6
  20. 1,217 Claude Fable 5
  21. 1,215 Qwen 3.8 Max
  22. 1,213 Claude Fable 5.1
  23. 1,211 Gemini 4 Carbon
  24. 1,209 Grok 4.7
  25. 1,208 GPT-5.5
  26. 1,207 GLM-5.3-Flash
  27. 1,205 Muse Spark Open
  28. 1,204 Claude Opus 5
  29. 1,203 Gemini 3.6 Flash
  30. 1,194 Grok 4.5
  31. 1,194 Muse Spark 1.1
  32. 1,189 Gemini 3.5 Flash
  33. 1,185 GPT-6 Luna
  34. 1,184 GPT-5.6 Terra
  35. 1,181 Kimi K2.7 Code
  36. 1,177 Claude Opus 4.8
  37. 1,175 Claude Sonnet 5
  38. 1,171 Qwen3.7-Plus
  39. 1,169 Kimi K2.6
  40. 1,169 MiniMax-M3.1-Flash
  41. 1,159 GPT-5.6 Luna
  42. 1,158 Nex N2 Pro
  43. 1,154 Kimi K2.5
  44. 1,151 MiniMax-M3
  45. 1,151 Claude Haiku 5.5
  46. 1,137 Tencent Hy4
  47. 1,130 Gemini 3.1 Pro
  48. 1,114 Mistral Large 4
  49. 1,112 Step 3.7 Flash
  50. 1,110 GLM-5V-Turbo
  51. 1,071 Muse Glimmer 30B
  52. 1,071 Grok 4.3
  53. 1,044 Nemotron 3.5 Super VL

Bars start at 1,000.

  1. 1 GPT-6 Astra xhigh, max 1,304
  2. 2 Muse Ember 1,273
  3. 3 Gemini 4 Argon 1,271
  4. 4 Kimi K3 1,269
  5. 5 Claude Opus 5.5 1,267
  6. 6 Muse Wonder max 1,264
  7. 7 Muse Spark 1.3 xhigh 1,257
  8. 8 Kimi K3.1 Pro 1,254
  9. 9 Claude Fable 5.2 1,247
  10. 10 GPT-6 Sol 1,233
  11. 11 Claude Sonnet 5.5 1,232
  12. 12 GPT-5.6 Sol medium 1,231
  13. 13 GPT-6.1 Sol 1,231
  14. 14 Gemini 3.8 Flash 1,230
  15. 15 Gemini 3.7 Flash 1,227
  16. 16 GPT-6.1 Terra 1,227
  17. 17 DeepSeek V4.1 Flash 1017 1,226
  18. 18 Muse Spark 1.2 1,218
  19. 19 Grok 4.6 1,218
  20. 20 Claude Fable 5 1,217
  21. 21 Qwen 3.8 Max 1,215
  22. 22 Claude Fable 5.1 1,213
  23. 23 Gemini 4 Carbon 1,211
  24. 24 Grok 4.7 1,209
  25. 25 GPT-5.5 1,208
  26. 26 GLM-5.3-Flash 1,207
  27. 27 Muse Spark Open xhigh 1,205
  28. 28 Claude Opus 5 1,204
  29. 29 Gemini 3.6 Flash 1,203
  30. 30 Grok 4.5 1,194
  31. 31 Muse Spark 1.1 1,194
  32. 32 Gemini 3.5 Flash 1,189
  33. 33 GPT-6 Luna 1,185
  34. 34 GPT-5.6 Terra 1,184
  35. 35 Kimi K2.7 Code 1,181
  36. 36 Claude Opus 4.8 1,177
  37. 37 Claude Sonnet 5 1,175
  38. 38 Qwen3.7-Plus 1,171
  39. 39 Kimi K2.6 1,169
  40. 40 MiniMax-M3.1-Flash 1,169
  41. 41 GPT-5.6 Luna 1,159
  42. 42 Nex N2 Pro 1,158
  43. 43 Kimi K2.5 thinking 1,154
  44. 44 MiniMax-M3 1,151
  45. 45 Claude Haiku 5.5 1,151
  46. 46 Tencent Hy4 1,137
  47. 47 Gemini 3.1 Pro 1,130
  48. 48 Mistral Large 4 1,114
  49. 49 Step 3.7 Flash 1,112
  50. 50 GLM-5V-Turbo 1,110
  51. 51 Muse Glimmer 30B xhigh 1,071
  52. 52 Grok 4.3 1,071
  53. 53 Nemotron 3.5 Super VL 1,044

Source: Official DesignArena Image to HTML leaderboard

#EntryScoreSettingDateSource
1 GPT-6 Astra 1,304 xhigh, max Sep 2026 Official leaderboard
2 Muse Ember 1,273 Nov 2026 Official leaderboard
3 Gemini 4 Argon 1,271 Sep 2026 Official leaderboard
4 Kimi K3 1,269 Jul 2026 Official leaderboard
5 Claude Opus 5.5 1,267 Sep 2026 Official leaderboard
6 Muse Wonder 1,264 max Oct 2026 Official leaderboard
7 Muse Spark 1.3 1,257 xhigh Sep 2026 Official leaderboard
8 Kimi K3.1 Pro 1,254 Oct 2026 Official leaderboard
9 Claude Fable 5.2 1,247 Nov 2026 Official leaderboard
10 GPT-6 Sol 1,233 Sep 2026 Official leaderboard
11 Claude Sonnet 5.5 1,232 Sep 2026 Official leaderboard
12 GPT-5.6 Sol 1,231 medium Jul 2026 Official leaderboard
13 GPT-6.1 Sol 1,231 Sep 2026 Official leaderboard
14 Gemini 3.8 Flash 1,230 Sep 2026 Official leaderboard
15 Gemini 3.7 Flash 1,227 Aug 2026 Official leaderboard
16 GPT-6.1 Terra 1,227 Oct 2026 Official leaderboard
17 DeepSeek V4.1 Flash 1,226 1017 Sep 2026 Official leaderboard
18 Muse Spark 1.2 1,218 Aug 2026 Official leaderboard
19 Grok 4.6 1,218 Aug 2026 Official leaderboard
20 Claude Fable 5 1,217 Jun 2026 Official leaderboard
21 Qwen 3.8 Max 1,215 Aug 2026 Official leaderboard
22 Claude Fable 5.1 1,213 Sep 2026 Official leaderboard
23 Gemini 4 Carbon 1,211 Oct 2026 Official leaderboard
24 Grok 4.7 1,209 Sep 2026 Official leaderboard
25 GPT-5.5 1,208 Apr 2026 Official leaderboard
26 GLM-5.3-Flash 1,207 Aug 2026 Official leaderboard
27 Muse Spark Open 1,205 xhigh Oct 2026 Official leaderboard
28 Claude Opus 5 1,204 Jul 2026 Official leaderboard
29 Gemini 3.6 Flash 1,203 Jul 2026 Official leaderboard
30 Grok 4.5 1,194 Jul 2026 Official leaderboard
31 Muse Spark 1.1 1,194 Jul 2026 Official leaderboard
32 Gemini 3.5 Flash 1,189 May 2026 Official leaderboard
33 GPT-6 Luna 1,185 Sep 2026 Official leaderboard
34 GPT-5.6 Terra 1,184 Jul 2026 Official leaderboard
35 Kimi K2.7 Code 1,181 Jun 2026 Official leaderboard
36 Claude Opus 4.8 1,177 May 2026 Official leaderboard
37 Claude Sonnet 5 1,175 Jun 2026 Official leaderboard
38 Qwen3.7-Plus 1,171 Jun 2026 Official leaderboard
39 Kimi K2.6 1,169 Apr 2026 Official leaderboard
40 MiniMax-M3.1-Flash 1,169 Oct 2026 Official leaderboard
41 GPT-5.6 Luna 1,159 Jul 2026 Official leaderboard
42 Nex N2 Pro 1,158 Official leaderboard
43 Kimi K2.5 1,154 thinking Jan 2026 Official leaderboard
44 MiniMax-M3 1,151 Jun 2026 Official leaderboard
45 Claude Haiku 5.5 1,151 Oct 2026 Official leaderboard
46 Tencent Hy4 1,137 Nov 2026 Official leaderboard
47 Gemini 3.1 Pro 1,130 Feb 2026 Official leaderboard
48 Mistral Large 4 1,114 Oct 2026 Official leaderboard
49 Step 3.7 Flash 1,112 May 2026 Official leaderboard
50 GLM-5V-Turbo 1,110 Apr 2026 Official leaderboard
51 Muse Glimmer 30B 1,071 xhigh Official leaderboard
52 Grok 4.3 1,071 Apr 2026 Official leaderboard
53 Nemotron 3.5 Super VL 1,044 Oct 2026 Official leaderboard

More effort adds a median 30 elo

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
GPT-6 Astra
xhigh xhigh, max
1,304
GPT-5.6 Sol
medium, xhi…
1,231
DeepSeek V4.1 Flash
1017
1,226

Source: Official DesignArena Image to HTML leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-6 Astra xhigh, max 1,304
GPT-6 Astra xhigh 1,274
Muse Ember 1,273
Gemini 4 Argon 1,271
Kimi K3 1,269
Claude Opus 5.5 1,267
Muse Wonder max 1,264
Muse Spark 1.3 xhigh 1,257
Kimi K3.1 Pro 1,254
Claude Fable 5.2 1,247
GPT-6 Sol 1,233
Claude Sonnet 5.5 1,232
GPT-5.6 Sol medium 1,231
GPT-6.1 Sol 1,231
Gemini 3.8 Flash 1,230
Gemini 3.7 Flash 1,227
GPT-6.1 Terra 1,227
DeepSeek V4.1 Flash 1017 1,226
GPT-5.6 Sol medium, xhigh 1,223
Muse Spark 1.2 1,218
Grok 4.6 1,218
Claude Fable 5 1,217
Qwen 3.8 Max 1,215
Claude Fable 5.1 1,213
Gemini 4 Carbon 1,211
Grok 4.7 1,209
GPT-5.5 1,208
GLM-5.3-Flash 1,207
DeepSeek V4.1 Flash 1,206
Muse Spark Open xhigh 1,205
Claude Opus 5 1,204
Gemini 3.6 Flash 1,203
Grok 4.5 1,194
Muse Spark 1.1 1,194
Gemini 3.5 Flash 1,189
GPT-6 Luna 1,185
GPT-5.6 Terra 1,184
Kimi K2.7 Code 1,181
Claude Opus 4.8 1,177
Claude Sonnet 5 1,175
Qwen3.7-Plus 1,171
Kimi K2.6 1,169
MiniMax-M3.1-Flash 1,169
GPT-5.6 Luna 1,159
Kimi K2.5 thinking 1,154
MiniMax-M3 1,151
Claude Haiku 5.5 1,151
Tencent Hy4 1,137
Gemini 3.1 Pro 1,130
Mistral Large 4 1,114
Step 3.7 Flash 1,112
GLM-5V-Turbo 1,110
Grok 4.3 1,071
Nemotron 3.5 Super VL 1,044

7 models have held number one on the official leaderboard since 2023

Lab claims count from the day the card came out, independent runs from the model’s release.

Moonshot AI OpenAI Anthropic
Own leaderboard 7 leaders
GPT-6 Astra 1,304

Source: Official DesignArena Image to HTML leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard
1,000 1,100 1,200 1,300 1,400 2026 2027 1,000 1,100 1,200 1,300 1,400 2026 2027

Source: Official DesignArena Image to HTML leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

More in Coding

All 282 in Coding
  • SWE-bench Verified In use 500 human-validated GitHub issues from 12 Python repos. Agent writes a patch that must pass hidden tests. Claude Opus 4.7 83.5%
  • LiveCodeBench v6 In use Competitive programming problems collected after model cutoffs from LeetCode, AtCoder, Codeforces. Scored pass@1 on hidden tests. DeepSeek-V4-Pro 93.5%
  • SWE-bench Multilingual In use 300 GitHub-issue tasks across nine programming languages in SWE-bench format. Patch graded by repository tests. Claude Fable 5.2 96.6%
  • DeepSWE In use 113 original long-horizon engineering tasks across 91 repos and five languages, written from scratch. Hand-written verifiers accept alternative solutions. GPT-6 Astra 74.1%
  • SciCode In use Scientist-curated research coding problems split into subproblems across natural sciences. Code checked against test cases. Seed2.1 Pro 59.8%
  • Codeforces rating In use Model solves recent Codeforces contest problems. An Elo-style rating is computed from simulated contest performance. DeepSeek V4.1 Flash 3,471