Benchmark Almanac

Deep20Bench 1.1

GPT-6 Astra 13.7 leads since September 2026
Not in model cards

Models identify hidden people, characters, or objects by asking questions. The score averages the number of counted questions, with a fixed penalty for failed rounds, over three rounds for each of ten subjects.

Edition 1.1 changes the answer options and question limit from edition 1.0. Model-based answer checks can require replaying rounds. September is the earliest month of the opened edition 1.1 results. The small subject set limits generalization.

Released
September 2026
Kind
Puzzles and chess
Level
Everyday
Scored
Graded by a model
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Multi-turn
Environment
Twenty questions
Run by
Anyone
In views
Games
Strongest model tested
Claude Fable 5.2 #2 on the capability index
Last model added
November 2026
Scores refreshed
November 18th, 2026
26
Entries on its own leaderboard, 25 models we track

2 models share the top score of 13.7 2 models share the top score of 13.7

Best score per entry from the benchmark's own leaderboard, read on November 18th, 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 GPT-6 Astra high 13.7
  2. 2 Gemini 4 Argon high 13.7
  3. 3 Claude Fable 5.2 high 14.6
  4. 4 GPT-6.2 Sol high 14.7
  5. 5 Gemini 3.8 Flash high 14.9
  6. 6 Gemini 4 Hydrogen high 14.9
  7. 7 Claude Fable 5.1 high 15.1
  8. 8 GPT-6.1 Sol high 15.2
  9. 9 Claude Opus 5.5 high 16.4
  10. 10 Gemini 3.7 Flash high 16.6
  11. 11 Grok 4.6 high 17.1
  12. 12 GPT-6 Sol high 18.0
  13. 13 GPT-5.6 Sol high 18.1
  14. 14 Claude Opus 5 high 18.1
  15. 15 Claude Sonnet 5.5 high 18.2
  16. 16 Muse Spark 1.3 Contributor high 18.4
  17. 17 GPT-6.1 Terra high 18.8
  18. 18 Muse Wonder high 19.6
  19. 19 Claude Haiku 5.5 high 20.0
  20. 20 GPT-5.6 Luna high 21.0
  21. 21 GPT-6 Luna high 21.2
  22. 22 gpt-oss-120b high 22.2
  23. 23 Grok 4.7 high 22.2
  24. 24 DeepSeek V4.1 Flash 1017 high 22.7
  25. 25 GLM-5.3-Flash high 23.3
  26. 26 MiniMax-M3 high 29.2

Source: Official Deep20Bench 1.1 leaderboard

#EntryScore ± 95% CISettingReleasedRun onSource
1 GPT-6 Astra 13.7 ±1.2 high Sep 2026 Sep 2026 Official leaderboard
2 Gemini 4 Argon 13.7 ±1.3 high Sep 2026 Oct 2026 Official leaderboard
3 Claude Fable 5.2 14.6 ±0.5 high Nov 2026 Nov 2026 Official leaderboard
4 GPT-6.2 Sol 14.7 ±2.8 high Nov 2026 Nov 2026 Official leaderboard
5 Gemini 3.8 Flash 14.9 ±1.7 high Sep 2026 Sep 2026 Official leaderboard
6 Gemini 4 Hydrogen 14.9 ±2.8 high Nov 2026 Nov 2026 Official leaderboard
7 Claude Fable 5.1 15.1 ±1.3 high Sep 2026 Sep 2026 Official leaderboard
8 GPT-6.1 Sol 15.2 ±2.4 high Sep 2026 Sep 2026 Official leaderboard
9 Claude Opus 5.5 16.4 ±1.9 high Sep 2026 Nov 2026 Official leaderboard
10 Gemini 3.7 Flash 16.6 ±2.2 high Aug 2026 Sep 2026 Official leaderboard
11 Grok 4.6 17.1 ±1.6 high Aug 2026 Sep 2026 Official leaderboard
12 GPT-6 Sol 18.0 ±1.7 high Sep 2026 Sep 2026 Official leaderboard
13 GPT-5.6 Sol 18.1 ±3.7 high Jul 2026 Sep 2026 Official leaderboard
14 Claude Opus 5 18.1 ±1.6 high Jul 2026 Sep 2026 Official leaderboard
15 Claude Sonnet 5.5 18.2 ±2.7 high Sep 2026 Nov 2026 Official leaderboard
16 Muse Spark 1.3 Contributor 18.4 ±1.8 high Sep 2026 Official leaderboard
17 GPT-6.1 Terra 18.8 ±2.9 high Oct 2026 Oct 2026 Official leaderboard
18 Muse Wonder 19.6 ±2.1 high Oct 2026 Oct 2026 Official leaderboard
19 Claude Haiku 5.5 20.0 ±2.9 high Oct 2026 Oct 2026 Official leaderboard
20 GPT-5.6 Luna 21.0 ±2.5 high Jul 2026 Sep 2026 Official leaderboard
21 GPT-6 Luna 21.2 ±3 high Sep 2026 Sep 2026 Official leaderboard
22 gpt-oss-120b 22.2 ±2.7 high Aug 2025 Sep 2026 Official leaderboard
23 Grok 4.7 22.2 ±1.4 high Sep 2026 Sep 2026 Official leaderboard
24 DeepSeek V4.1 Flash 1017 22.7 ±3.1 high Oct 2026 Oct 2026 Official leaderboard
25 GLM-5.3-Flash 23.3 ±3.7 high Aug 2026 Sep 2026 Official leaderboard
26 MiniMax-M3 29.2 ±4.5 high Jun 2026 Sep 2026 Official leaderboard
#ModelScore ± 95% CISettingReleasedRun onSource
1 GPT-6 Astra 13.7 ±1.2 high Sep 2026 Sep 2026 Official leaderboard
2 Gemini 4 Argon 13.7 ±1.3 high Sep 2026 Oct 2026 Official leaderboard
3 Claude Fable 5.2 14.6 ±0.5 high Nov 2026 Nov 2026 Official leaderboard
4 GPT-6.2 Sol 14.7 ±2.8 high Nov 2026 Nov 2026 Official leaderboard
5 Gemini 3.8 Flash 14.9 ±1.7 high Sep 2026 Sep 2026 Official leaderboard
6 Gemini 4 Hydrogen 14.9 ±2.8 high Nov 2026 Nov 2026 Official leaderboard
7 Claude Fable 5.1 15.1 ±1.3 high Sep 2026 Sep 2026 Official leaderboard
8 GPT-6.1 Sol 15.2 ±2.4 high Sep 2026 Sep 2026 Official leaderboard
9 Claude Opus 5.5 16.4 ±1.9 high Sep 2026 Nov 2026 Official leaderboard
10 Gemini 3.7 Flash 16.6 ±2.2 high Aug 2026 Sep 2026 Official leaderboard
11 Grok 4.6 17.1 ±1.6 high Aug 2026 Sep 2026 Official leaderboard
12 GPT-6 Sol 18.0 ±1.7 high Sep 2026 Sep 2026 Official leaderboard
13 GPT-5.6 Sol 18.1 ±3.7 high Jul 2026 Sep 2026 Official leaderboard
14 Claude Opus 5 18.1 ±1.6 high Jul 2026 Sep 2026 Official leaderboard
15 Claude Sonnet 5.5 18.2 ±2.7 high Sep 2026 Nov 2026 Official leaderboard
16 GPT-6.1 Terra 18.8 ±2.9 high Oct 2026 Oct 2026 Official leaderboard
17 Muse Wonder 19.6 ±2.1 high Oct 2026 Oct 2026 Official leaderboard
18 Claude Haiku 5.5 20.0 ±2.9 high Oct 2026 Oct 2026 Official leaderboard
19 GPT-5.6 Luna 21.0 ±2.5 high Jul 2026 Sep 2026 Official leaderboard
20 GPT-6 Luna 21.2 ±3 high Sep 2026 Sep 2026 Official leaderboard
21 gpt-oss-120b 22.2 ±2.7 high Aug 2025 Sep 2026 Official leaderboard
22 Grok 4.7 22.2 ±1.4 high Sep 2026 Sep 2026 Official leaderboard
23 DeepSeek V4.1 Flash 1017 22.7 ±3.1 high Oct 2026 Oct 2026 Official leaderboard
24 GLM-5.3-Flash 23.3 ±3.7 high Aug 2026 Sep 2026 Official leaderboard
25 MiniMax-M3 29.2 ±4.5 high Jun 2026 Sep 2026 Official leaderboard

3 models have held number one on the official leaderboard since its release in September 2026

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

Anthropic Google DeepMind OpenAI
Own leaderboard 3 leaders

Source: Official Deep20Bench 1.1 leaderboard

Scores over time

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
10.0 15.0 20.0 25.0 30.0 2026 2027 10.0 15.0 20.0 25.0 30.0 2026 2027

Source: Official Deep20Bench 1.1 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

All 10 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (26 entries).
  • Pass Models scored 25 models we track, plus 1 other entry on its leaderboard.
  • Pass Newest score The newest is from November 2026, for GPT-6.2 Sol.
  • Pass Frontier models 8 of the top 10 have a score, led by Claude Fable 5.2 (#2).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on deep20bench.com: 25 rows for models we track and 1 other entry at the last read.
  • Pass Last read November 18th, 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