Benchmark Almanac

DesignArena Godot Game Dev

GPT-6 Astra 1,319 leads since September 2026
Not in model cards

Coding agents build complete Godot projects with GDScript, scenes and generated assets, exported to run in the browser. Users play the games and vote.

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

Kind
Building whole projects
Level
Everyday
Scored
Ranked by votes
Answers
Open-ended
Tasks
All public
Strongest model tested
GPT-6 Astra #3 on the capability index
Last model added
September 2026
Scores refreshed
November 12th, 2026
24
Models scored on its own leaderboard

GPT-6 Astra leads with 1,319 among models released by September 2026

Best score per model from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best.

The newest model with a score here came out in September 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

24 models. Scroll sideways for the rest. Bars start at 900.

  1. 1,319 GPT-6 Astra
  2. 1,307 Claude Fable 5.1
  3. 1,274 Claude Fable 5
  4. 1,261 Claude Opus 5
  5. 1,256 Grok 4.6
  6. 1,251 Gemini 3.8 Flash
  7. 1,239 GPT-5.6 Sol
  8. 1,239 Gemini 3.7 Flash
  9. 1,221 Muse Spark 1.3
  10. 1,215 Qwen 3.8 Max
  11. 1,214 Claude Opus 4.8
  12. 1,209 Grok 4.5
  13. 1,205 Claude Sonnet 5
  14. 1,205 Kimi K3
  15. 1,169 Qwen3.7-Max
  16. 1,155 GPT-5.5
  17. 1,145 Gemini 3.1 Pro
  18. 1,128 Kimi K2.7 Code
  19. 1,103 Kimi K2.6
  20. 1,086 Muse Spark 1.1
  21. 1,081 Gemini 3.5 Flash
  22. 1,039 GLM-5.1
  23. 940 GLM-5V-Turbo
  24. 919 Grok 4.3

Bars start at 900.

  1. 1 GPT-6 Astra xhigh 1,319
  2. 2 Claude Fable 5.1 1,307
  3. 3 Claude Fable 5 1,274
  4. 4 Claude Opus 5 1,261
  5. 5 Grok 4.6 1,256
  6. 6 Gemini 3.8 Flash 1,251
  7. 7 GPT-5.6 Sol medium, xhigh 1,239
  8. 8 Gemini 3.7 Flash 1,239
  9. 9 Muse Spark 1.3 xhigh, max 1,221
  10. 10 Qwen 3.8 Max 1,215
  11. 11 Claude Opus 4.8 1,214
  12. 12 Grok 4.5 1,209
  13. 13 Claude Sonnet 5 1,205
  14. 14 Kimi K3 1,205
  15. 15 Qwen3.7-Max 1,169
  16. 16 GPT-5.5 1,155
  17. 17 Gemini 3.1 Pro 1,145
  18. 18 Kimi K2.7 Code 1,128
  19. 19 Kimi K2.6 1,103
  20. 20 Muse Spark 1.1 1,086
  21. 21 Gemini 3.5 Flash 1,081
  22. 22 GLM-5.1 1,039
  23. 23 GLM-5V-Turbo 940
  24. 24 Grok 4.3 919

Source: Official DesignArena Godot Game Dev leaderboard

#ModelScoreSettingDateSource
1 GPT-6 Astra 1,319 xhigh Sep 2026 Official leaderboard
2 Claude Fable 5.1 1,307 Sep 2026 Official leaderboard
3 Claude Fable 5 1,274 Jun 2026 Official leaderboard
4 Claude Opus 5 1,261 Jul 2026 Official leaderboard
5 Grok 4.6 1,256 Aug 2026 Official leaderboard
6 Gemini 3.8 Flash 1,251 Sep 2026 Official leaderboard
7 GPT-5.6 Sol 1,239 medium, xhigh Jul 2026 Official leaderboard
8 Gemini 3.7 Flash 1,239 Aug 2026 Official leaderboard
9 Muse Spark 1.3 1,221 xhigh, max Sep 2026 Official leaderboard
10 Qwen 3.8 Max 1,215 Aug 2026 Official leaderboard
11 Claude Opus 4.8 1,214 May 2026 Official leaderboard
12 Grok 4.5 1,209 Jul 2026 Official leaderboard
13 Claude Sonnet 5 1,205 Jun 2026 Official leaderboard
14 Kimi K3 1,205 Jul 2026 Official leaderboard
15 Qwen3.7-Max 1,169 May 2026 Official leaderboard
16 GPT-5.5 1,155 Apr 2026 Official leaderboard
17 Gemini 3.1 Pro 1,145 Feb 2026 Official leaderboard
18 Kimi K2.7 Code 1,128 Jun 2026 Official leaderboard
19 Kimi K2.6 1,103 Apr 2026 Official leaderboard
20 Muse Spark 1.1 1,086 Jul 2026 Official leaderboard
21 Gemini 3.5 Flash 1,081 May 2026 Official leaderboard
22 GLM-5.1 1,039 Apr 2026 Official leaderboard
23 GLM-5V-Turbo 940 Apr 2026 Official leaderboard
24 Grok 4.3 919 Apr 2026 Official leaderboard

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.

Google DeepMind OpenAI Qwen Anthropic
Own leaderboard Not tested since Sep 2026
GPT-6 Astra 1,319

Source: Official DesignArena Godot Game Dev 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
900 1,000 1,100 1,200 1,300 1,400 2026 900 1,000 1,100 1,200 1,300 1,400 2026

Source: Official DesignArena Godot Game Dev 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