Benchmark Almanac

Game Arena Ultimate Tic-Tac-Toe

GPT-6 Astra 608 leads since Sep 2026 Not in model cards

Language models play nine linked tic-tac-toe boards. Each move decides which small board the opponent must play next.

Kaggle published the leaderboard on 2026-07-28 with 36 models on 2026-11-01. Gameplay logs are a public Kaggle dataset. Part of the Kaggle Game Arena suite.

Made by Kaggle and Google DeepMind Released Jul 2026 Puzzles and chess All tasks public Last model added Oct 2026 Scores refreshed 2 Nov 2026
Level
Everyday Most adults, without training
Scored
Ranked by matches Models play against each other, and the results become ratings.
Answers
Agent task Many steps in a terminal, browser, desktop, app, game or robot.
Kaggle/kaggle-environments 459 stars Dataset Leaderboard kaggle.com
36
Models scored on its own leaderboard

GPT-6 Astra leads with 608

Best score per model from the benchmark's own leaderboard, read on 2 Nov 2026. Where a model ran with several settings or harnesses, we keep its best.

36 models. Scroll sideways for the rest.

  1. 608 GPT-6 Astra
  2. 591 Claude Opus 5
  3. 544 Grok 4.7
  4. 505 Grok 4.6
  5. 497 GPT-6.1 Sol
  6. 486 GPT-5.5
  7. 484 Gemini 4 Argon
  8. 449 GPT-5.6 Sol
  9. 442 Claude Fable 5.1
  10. 426 Claude Opus 4.8
  11. 392 GPT-6 Sol
  12. 381 Grok 4.5
  13. 380 Gemini 3.8 Flash
  14. 369 GPT-5.4
  15. 363 Claude Opus 5.5
  16. 359 Gemini 3.1 Pro
  17. 286 Claude Opus 4.6
  18. 271 Gemini 3.6 Flash
  19. 269 Gemini 3.7 Flash
  20. 263 Gemini 3.5 Flash
  21. 247 Claude Sonnet 5.5
  22. 244 GPT-5.2
  23. 243 o3
  24. 188 Claude Sonnet 5
  25. 188 Claude Sonnet 4.6
  26. 178 Claude Sonnet 4.5
  27. 169 Grok 4.20
  28. 169 Gemini 3 Flash
  29. 85 Claude Haiku 5.5
  30. 57 GPT-5.6 Terra
  31. 57 Claude Opus 4.7
  32. 27 Grok 4.3
  33. 22 GPT-6 Luna
  34. 21 GPT-5.6 Luna
  35. 17 GPT-5 mini
  36. 0 GPT-5.4 Mini
  1. 1 GPT-6 Astra 608
  2. 2 Claude Opus 5 591
  3. 3 Grok 4.7 544
  4. 4 Grok 4.6 505
  5. 5 GPT-6.1 Sol 497
  6. 6 GPT-5.5 486
  7. 7 Gemini 4 Argon 484
  8. 8 GPT-5.6 Sol 449
  9. 9 Claude Fable 5.1 442
  10. 10 Claude Opus 4.8 426
  11. 11 GPT-6 Sol 392
  12. 12 Grok 4.5 381
  13. 13 Gemini 3.8 Flash 380
  14. 14 GPT-5.4 369
  15. 15 Claude Opus 5.5 363
  16. 16 Gemini 3.1 Pro 359
  17. 17 Claude Opus 4.6 286
  18. 18 Gemini 3.6 Flash 271
  19. 19 Gemini 3.7 Flash 269
  20. 20 Gemini 3.5 Flash 263
  21. 21 Claude Sonnet 5.5 247
  22. 22 GPT-5.2 244
  23. 23 o3 243
  24. 24 Claude Sonnet 5 188
  25. 25 Claude Sonnet 4.6 188
  26. 26 Claude Sonnet 4.5 178
  27. 27 Grok 4.20 reasoning 169
  28. 28 Gemini 3 Flash 169
  29. 29 Claude Haiku 5.5 85
  30. 30 GPT-5.6 Terra 57
  31. 31 Claude Opus 4.7 57
  32. 32 Grok 4.3 27
  33. 33 GPT-6 Luna 22
  34. 34 GPT-5.6 Luna 21
  35. 35 GPT-5 mini 17
  36. 36 GPT-5.4 Mini 0
#ModelScoreSettingDateSource
1 GPT-6 Astra 608 Sep 2026 Official leaderboard
2 Claude Opus 5 591 Jul 2026 Official leaderboard
3 Grok 4.7 544 Sep 2026 Official leaderboard
4 Grok 4.6 505 Aug 2026 Official leaderboard
5 GPT-6.1 Sol 497 Sep 2026 Official leaderboard
6 GPT-5.5 486 Apr 2026 Official leaderboard
7 Gemini 4 Argon 484 Sep 2026 Official leaderboard
8 GPT-5.6 Sol 449 Jul 2026 Official leaderboard
9 Claude Fable 5.1 442 Sep 2026 Official leaderboard
10 Claude Opus 4.8 426 May 2026 Official leaderboard
11 GPT-6 Sol 392 Sep 2026 Official leaderboard
12 Grok 4.5 381 Jul 2026 Official leaderboard
13 Gemini 3.8 Flash 380 Sep 2026 Official leaderboard
14 GPT-5.4 369 Mar 2026 Official leaderboard
15 Claude Opus 5.5 363 Sep 2026 Official leaderboard
16 Gemini 3.1 Pro 359 Feb 2026 Official leaderboard
17 Claude Opus 4.6 286 Feb 2026 Official leaderboard
18 Gemini 3.6 Flash 271 Jul 2026 Official leaderboard
19 Gemini 3.7 Flash 269 Aug 2026 Official leaderboard
20 Gemini 3.5 Flash 263 May 2026 Official leaderboard
21 Claude Sonnet 5.5 247 Sep 2026 Official leaderboard
22 GPT-5.2 244 Dec 2025 Official leaderboard
23 o3 243 Apr 2025 Official leaderboard
24 Claude Sonnet 5 188 Jun 2026 Official leaderboard
25 Claude Sonnet 4.6 188 Feb 2026 Official leaderboard
26 Claude Sonnet 4.5 178 Sep 2025 Official leaderboard
27 Grok 4.20 169 reasoning Feb 2026 Official leaderboard
28 Gemini 3 Flash 169 Dec 2025 Official leaderboard
29 Claude Haiku 5.5 85 Oct 2026 Official leaderboard
30 GPT-5.6 Terra 57 Jul 2026 Official leaderboard
31 Claude Opus 4.7 57 Apr 2026 Official leaderboard
32 Grok 4.3 27 Apr 2026 Official leaderboard
33 GPT-6 Luna 22 Sep 2026 Official leaderboard
34 GPT-5.6 Luna 21 Jul 2026 Official leaderboard
35 GPT-5 mini 17 Aug 2025 Official leaderboard
36 GPT-5.4 Mini 0 Mar 2026 Official leaderboard

3 models have held number one since its release in Jul 2026

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

OpenAI Anthropic
Own leaderboard 3 leaders
GPT-6 Astra 608

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
0 200 400 600 800 2025 2026 0 200 400 600 800 2025 2026

No lab model card reports it yet

We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark.

More in Games and long tasks

All 68 in Games and long tasks
  • ARC-AGI-3 In use Interactive game environments where agents must explore, infer goals and rules, and act efficiently without instructions. GPT-6 Astra 100%
  • MineDojo Fading Agents complete open-ended tasks in Minecraft, such as mining blocks and fighting mobs, from a large simulation suite. Programmatic checks score many tasks. Seed2.0 Pro 49%
  • Orak Fading LLM agents play 12 popular video games across most genres through an MCP interface. Scores come from in-game progress or outcomes per title. Nemotron 3 Nano 4B 22.9
  • GameWorld Only in system cards Multimodal game agents play 34 browser games across 170 tasks. Outcomes are checked from the game state. Gemini 3 Flash 41.9%
  • Swarm Bench Only in system cards Models orchestrate agent swarms on complex tasks that benefit from coordinated decomposition and parallel execution. Kimi K3 76.3%
  • TextQuests Only in system cards Agents play 25 classic interactive fiction games. Measures game progress and harmful actions per 100 steps. Muse Spark 2.40