Benchmark Almanac

The Aggregate Puzzle Portfolio Ladder

GPT-5.2 96.7% leads since December 2025
Not in model cards

Models solve 30 generated puzzles across five families. Each family has six questions with answers fixed by the generator.

The maker tracks 29 models. Some models answered fewer than 30 questions, and the reported accuracy uses their own item count.

Kind
Puzzles and chess
Level
University
Scored
Checked automatically
Answers
Short answer
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Single turn
Items from
Generated
Refresh
One fixed set
Run by
A host
Strongest model tested
GPT-6.2 Astra #1 on the capability index
Last model added
November 2026
Scores refreshed
November 20th, 2026
29
Entries on its own leaderboard, 24 models we track

6 models share the top score of 96.7% 6 models share the top score of 96.7%

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

  1. 1 Particle-1 free route, unset 96.7%
  2. 2 GPT-5.2 unset 96.7%
  3. 3 GPT-5.5 unset 96.7%
  4. 4 MiniMax-M3.1-Flash free route, unset 96.7%
  5. 5 Claude Fable 5.2 unset 96.7%
  6. 6 GPT-6.2 Astra unset 96.7%
  7. 7 GPT-6.2 Sol unset 96.7%
  8. 8 Ling-3.1-flash unset 96.3%
  9. 9 GLM-5.3 Mistral API, unset 92.9%
  10. 10 Ling-3.2-flash free route, unset 92.6%
  11. 11 Nemotron 3.5 Super VL free route, unset 92.6%
  12. 12 Kimi K2.6 unset 90%
  13. 13 Claude Opus 4.8 unset 90%
  14. 14 Qwen 3.6 Plus unset 86.7%
  15. 15 DeepSeek-V4-Pro unset 86.7%
  16. 16 ERNIE 5.5 Flash Preview free route, unset 86.2%
  17. 17 gpt-oss-120b unset 83.3%
  18. 18 Nex N2.5 Pro free route, unset 82.1%
  19. 19 Nemotron 3 Ultra free route, unset 77.8%
  20. 20 DeepSeek V4.1 Flash 1017 Mistral API, unset 76.7%
  21. 21 Gemma 4 31B IT free route, unset 73.3%
  22. 22 Flint free route, unset 66.7%
  23. 23 Nemotron 3 Super free route, unset 66.7%
  24. 24 Ling-3.0-flash-VL free route, unset 63.3%
  25. 25 Space Bunny Alpha stealth release, unset 60%
  26. 26 GPT-5.4 unset 60%
  27. 27 Gemma 4 26B A4B free route, unset 53.3%
  28. 28 Laguna XS 2.1 free route, unset 44.8%
  29. 29 Kimi K2 Sep 2025, unset 43.3%

Source: Official The Aggregate Puzzle Portfolio Ladder leaderboard

#EntryScoreSettingDateSource
1 Particle-1 96.7% free route, unset Official leaderboard
2 GPT-5.2 96.7% unset Dec 2025 Official leaderboard
3 GPT-5.5 96.7% unset Apr 2026 Official leaderboard
4 MiniMax-M3.1-Flash 96.7% free route, unset Oct 2026 Official leaderboard
5 Claude Fable 5.2 96.7% unset Nov 2026 Official leaderboard
6 GPT-6.2 Astra 96.7% unset Nov 2026 Official leaderboard
7 GPT-6.2 Sol 96.7% unset Nov 2026 Official leaderboard
8 Ling-3.1-flash 96.3% unset Oct 2026 Official leaderboard
9 GLM-5.3 92.9% Mistral API, unset Aug 2026 Official leaderboard
10 Ling-3.2-flash 92.6% free route, unset Official leaderboard
11 Nemotron 3.5 Super VL 92.6% free route, unset Oct 2026 Official leaderboard
12 Kimi K2.6 90% unset Apr 2026 Official leaderboard
13 Claude Opus 4.8 90% unset May 2026 Official leaderboard
14 Qwen 3.6 Plus 86.7% unset Mar 2026 Official leaderboard
15 DeepSeek-V4-Pro 86.7% unset Apr 2026 Official leaderboard
16 ERNIE 5.5 Flash Preview 86.2% free route, unset Official leaderboard
17 gpt-oss-120b 83.3% unset Aug 2025 Official leaderboard
18 Nex N2.5 Pro 82.1% free route, unset Official leaderboard
19 Nemotron 3 Ultra 77.8% free route, unset Jun 2026 Official leaderboard
20 DeepSeek V4.1 Flash 1017 76.7% Mistral API, unset Oct 2026 Official leaderboard
21 Gemma 4 31B IT 73.3% free route, unset Apr 2026 Official leaderboard
22 Flint 66.7% free route, unset Official leaderboard
23 Nemotron 3 Super 66.7% free route, unset Mar 2026 Official leaderboard
24 Ling-3.0-flash-VL 63.3% free route, unset Official leaderboard
25 Space Bunny Alpha 60% stealth release, unset Official leaderboard
26 GPT-5.4 60% unset Mar 2026 Official leaderboard
27 Gemma 4 26B A4B 53.3% free route, unset Apr 2026 Official leaderboard
28 Laguna XS 2.1 44.8% free route, unset Jul 2026 Official leaderboard
29 Kimi K2 Sep 2025 43.3% unset Sep 2025 Official leaderboard
#ModelScoreSettingDateSource
1 GPT-5.2 96.7% unset Dec 2025 Official leaderboard
2 GPT-5.5 96.7% unset Apr 2026 Official leaderboard
3 MiniMax-M3.1-Flash 96.7% free route, unset Oct 2026 Official leaderboard
4 Claude Fable 5.2 96.7% unset Nov 2026 Official leaderboard
5 GPT-6.2 Astra 96.7% unset Nov 2026 Official leaderboard
6 GPT-6.2 Sol 96.7% unset Nov 2026 Official leaderboard
7 Ling-3.1-flash 96.3% unset Oct 2026 Official leaderboard
8 GLM-5.3 92.9% Mistral API, unset Aug 2026 Official leaderboard
9 Ling-3.2-flash 92.6% free route, unset Official leaderboard
10 Nemotron 3.5 Super VL 92.6% free route, unset Oct 2026 Official leaderboard
11 Kimi K2.6 90% unset Apr 2026 Official leaderboard
12 Claude Opus 4.8 90% unset May 2026 Official leaderboard
13 Qwen 3.6 Plus 86.7% unset Mar 2026 Official leaderboard
14 DeepSeek-V4-Pro 86.7% unset Apr 2026 Official leaderboard
15 gpt-oss-120b 83.3% unset Aug 2025 Official leaderboard
16 Nemotron 3 Ultra 77.8% free route, unset Jun 2026 Official leaderboard
17 DeepSeek V4.1 Flash 1017 76.7% Mistral API, unset Oct 2026 Official leaderboard
18 Gemma 4 31B IT 73.3% free route, unset Apr 2026 Official leaderboard
19 Flint 66.7% free route, unset Official leaderboard
20 Nemotron 3 Super 66.7% free route, unset Mar 2026 Official leaderboard
21 GPT-5.4 60% unset Mar 2026 Official leaderboard
22 Gemma 4 26B A4B 53.3% free route, unset Apr 2026 Official leaderboard
23 Laguna XS 2.1 44.8% free route, unset Jul 2026 Official leaderboard
24 Kimi K2 Sep 2025 43.3% unset Sep 2025 Official leaderboard

2 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.

OpenAI
Own leaderboard Passed 85% in Dec 2025, by GPT-5.2
GPT-5.2 96.7%

Source: Official The Aggregate Puzzle Portfolio Ladder leaderboard

GPT-5.2 on the official leaderboard passed 85% in December 2025

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

Plot with other benchmarks
On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) GPT-5.2 gpt-oss-120b 0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) GPT-5.2 gpt-oss-120b

Source: Official The Aggregate Puzzle Portfolio Ladder leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Games and long tasks

All 261 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-6.2 Astra 100%
  • Orak Fading LLM agents play 12 popular video games across most genres through an MCP interface. Each game scores in-game progress or outcomes, and the board ranks models by their average rank across the games. Gemini 2.5 Pro 3.90
  • 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%
  • TextQuests Only in system cards Agents play 25 classic interactive fiction games. The score is harmful actions per 100 steps, where lower is better. The board also reports game progress. Llama 4 Scout 7.00
  • GameWorld Only in system cards Multimodal game agents play 34 browser games across 170 tasks. Outcomes are checked from the game state. GPT-6.1 Terra 76.8%
  • 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%

Data audit

8 of 9 checks pass, 1 warning. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark

Catalog record 1 warning 1 pass
  • Warning Task access We don’t know whether the tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (29 entries).
  • Pass Models scored 24 models we track, plus 5 other entries on its leaderboard.
  • Pass Newest score The newest is from November 2026, for GPT-6.2 Astra.
  • Note Frontier models 2 of the top 10 have a score, which is fine for a retired test. The strongest tested is GPT-6.2 Astra (#1).
  • Note Headroom Saturated. The best is 96.7%, past 85% since December 2025.
Leaderboard import 3 pass
  • Pass Importer We read the board on theaggregate.ai: 24 rows for models we track and 5 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check