Benchmark Almanac

Crest Arcade Chess

GPT-6.2 Astra 1,758 leads since November 2026
Not in model cards

Models play standard chess and Chess960 against fixed engine opponents without a list of legal moves. The organizer fits separate ratings for each variant.

Glimmer Index reports both variants for 20 distinct models. Each model encounters the same Chess960 starting positions in order, with alternating colors.

Made by
Glimmer Index
Kind
Multi-agent games
Level
School
Scored
Ranked by matches
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Actions
Interaction
Multi-agent
Environment
Chess
Refresh
One fixed set
Run by
A host
In views
Games, Agents
Strongest model tested
GPT-6.2 Astra #1 on the capability index
Last model added
November 2026
Scores refreshed
November 20th, 2026
20
Entries on its own leaderboard, 19 models we track

GPT-6.2 Astra leads with 1,758 GPT-6.2 Astra leads with 1,758

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. 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. This view shows only models we track, each at its best.

  1. 1 GPT-6.2 Astra medium 1,758
  2. 2 Gemini 4 Hydrogen medium 1,557
  3. 3 GPT-6.1 Sol medium 1,508
  4. 4 Gemini 4 Neon medium 1,497
  5. 5 Gemini 4 Argon medium 1,489
  6. 6 Gemini 4 Carbon medium 1,489
  7. 7 GPT-6 Astra medium 1,429
  8. 8 GPT-6 Sol medium 1,410
  9. 9 Claude Opus 5.5 medium 1,346
  10. 10 Claude Sonnet 5.5 medium 1,163
  11. 11 Claude Haiku 5.5 medium 1,090
  12. 12 GPT-6.1 Terra medium 1,085
  13. 13 Claude Fable 5.1 medium 1,083
  14. 14 gpt-oss-20b low 1,062
  15. 15 GPT-6 Luna medium 761
  16. 16 Qwen3.8-Flash-Next low 502
  17. 17 Gemma 3 4B none 99
  18. 18 Qwen3.5-9B none 62
  19. 19 Qwen 3.8 27B none 62
  20. 20 Ministral 3 8B none 62

Source: Official Crest Arcade Chess leaderboard

#EntryScore ±SettingReleasedRun onSource
1 GPT-6.2 Astra 1,758 ±120.5 medium Nov 2026 Nov 2026 Official leaderboard
2 Gemini 4 Hydrogen 1,557 ±126 medium Nov 2026 Nov 2026 Official leaderboard
3 GPT-6.1 Sol 1,508 ±119 medium Sep 2026 Oct 2026 Official leaderboard
4 Gemini 4 Neon 1,497 ±404.5 medium Nov 2026 Nov 2026 Official leaderboard
5 Gemini 4 Argon 1,489 ±737 medium Sep 2026 Nov 2026 Official leaderboard
6 Gemini 4 Carbon 1,489 ±737 medium Oct 2026 Nov 2026 Official leaderboard
7 GPT-6 Astra 1,429 ±120 medium Sep 2026 Oct 2026 Official leaderboard
8 GPT-6 Sol 1,410 ±131.5 medium Sep 2026 Oct 2026 Official leaderboard
9 Claude Opus 5.5 1,346 ±118.5 medium Sep 2026 Oct 2026 Official leaderboard
10 Claude Sonnet 5.5 1,163 ±117 medium Sep 2026 Oct 2026 Official leaderboard
11 Claude Haiku 5.5 1,090 ±115.5 medium Oct 2026 Oct 2026 Official leaderboard
12 GPT-6.1 Terra 1,085 ±113.5 medium Oct 2026 Oct 2026 Official leaderboard
13 Claude Fable 5.1 1,083 ±116 medium Sep 2026 Oct 2026 Official leaderboard
14 gpt-oss-20b 1,062 ±446.5 low Aug 2025 Oct 2026 Official leaderboard
15 GPT-6 Luna 761 ±116 medium Sep 2026 Oct 2026 Official leaderboard
16 Qwen3.8-Flash-Next 502 ±178.5 low Aug 2026 Oct 2026 Official leaderboard
17 Gemma 3 4B 99 ±464.5 none Mar 2025 Oct 2026 Official leaderboard
18 Qwen3.5-9B 62 ±454.5 none Feb 2026 Oct 2026 Official leaderboard
19 Qwen 3.8 27B 62 ±454.5 none Aug 2026 Oct 2026 Official leaderboard
20 Ministral 3 8B open weights 62 ±454.5 none Oct 2026 Official leaderboard
#ModelScore ±SettingReleasedRun onSource
1 GPT-6.2 Astra 1,758 ±120.5 medium Nov 2026 Nov 2026 Official leaderboard
2 Gemini 4 Hydrogen 1,557 ±126 medium Nov 2026 Nov 2026 Official leaderboard
3 GPT-6.1 Sol 1,508 ±119 medium Sep 2026 Oct 2026 Official leaderboard
4 Gemini 4 Neon 1,497 ±404.5 medium Nov 2026 Nov 2026 Official leaderboard
5 Gemini 4 Argon 1,489 ±737 medium Sep 2026 Nov 2026 Official leaderboard
6 Gemini 4 Carbon 1,489 ±737 medium Oct 2026 Nov 2026 Official leaderboard
7 GPT-6 Astra 1,429 ±120 medium Sep 2026 Oct 2026 Official leaderboard
8 GPT-6 Sol 1,410 ±131.5 medium Sep 2026 Oct 2026 Official leaderboard
9 Claude Opus 5.5 1,346 ±118.5 medium Sep 2026 Oct 2026 Official leaderboard
10 Claude Sonnet 5.5 1,163 ±117 medium Sep 2026 Oct 2026 Official leaderboard
11 Claude Haiku 5.5 1,090 ±115.5 medium Oct 2026 Oct 2026 Official leaderboard
12 GPT-6.1 Terra 1,085 ±113.5 medium Oct 2026 Oct 2026 Official leaderboard
13 Claude Fable 5.1 1,083 ±116 medium Sep 2026 Oct 2026 Official leaderboard
14 gpt-oss-20b 1,062 ±446.5 low Aug 2025 Oct 2026 Official leaderboard
15 GPT-6 Luna 761 ±116 medium Sep 2026 Oct 2026 Official leaderboard
16 Qwen3.8-Flash-Next 502 ±178.5 low Aug 2026 Oct 2026 Official leaderboard
17 Gemma 3 4B 99 ±464.5 none Mar 2025 Oct 2026 Official leaderboard
18 Qwen3.5-9B 62 ±454.5 none Feb 2026 Oct 2026 Official leaderboard
19 Qwen 3.8 27B 62 ±454.5 none Aug 2026 Oct 2026 Official leaderboard

Rankings by part

We read 2 parts from the benchmark’s own leaderboard: Standard and Chess960. Each ranks models on one part of the test. GPT-6.2 Astra leads both. Arrows show how far a model moved from its overall rank among the models we track.

Standard

19 models we track have a score here.

  1. 1 GPT-6.2 Astra 1,768
  2. 2 Gemini 4 Hydrogen 1,749
  3. 3 Gemini 4 Neon ↑1 1,713
  4. 4 GPT-6 Sol ↑4 1,673
  5. 5 GPT-6 Astra ↑2 1,561
  6. 6 GPT-6.1 Sol ↓3 1,481
  7. 7 Claude Opus 5.5 ↑2 1,456
  8. 8 gpt-oss-20b ↑6 1,414
  9. 9 Gemini 4 Argon ↓4 1,414
  10. 10 Gemini 4 Carbon ↓4 1,414

The best score, 1,758, costs $1.37 a task

Runs on the cost frontier score higher than any run that's cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

OpenAI Google DeepMind Anthropic Cost frontier One model at each effort level
0 500 1,000 1,500 2,000 $0.01 $0.10 $1.00 $10 Cost per task, log scale GPT-6.2 Astra, medium Gemini 4 Hydrogen, medium Gemini 4 Argon, medium Claude Haiku 5.5, medium GPT-6 Luna, medium 0 500 1,000 1,500 2,000 $0.01 $0.10 $1.00 $10 Cost per task, log scale GPT-6.2 Astra, medium Gemini 4 Hydrogen, medium

Source: Official Crest Arcade Chess leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-6.2 Astra medium 1,758 $1.37
Gemini 4 Hydrogen medium 1,557 $1.18
GPT-6.1 Sol medium 1,508 $0.90
Gemini 4 Neon medium 1,497 $2.19
Gemini 4 Argon medium 1,489 $0.30
Gemini 4 Carbon medium 1,489 $0.87
GPT-6 Astra medium 1,429 $2.38
GPT-6 Sol medium 1,410 $0.88
Claude Opus 5.5 medium 1,346 $0.91
Claude Sonnet 5.5 medium 1,163 $0.47
Claude Haiku 5.5 medium 1,090 $0.05
GPT-6.1 Terra medium 1,085 $0.11
Claude Fable 5.1 medium 1,083 $2.90
gpt-oss-20b low 1,062
GPT-6 Luna medium 761 $0.02
Qwen3.8-Flash-Next low 502
Gemma 3 4B none 99
Qwen3.5-9B none 62
Qwen 3.8 27B none 62

The best score, 1,758, costs $1.37 a task

Runs on the cost frontier score higher than any run that's cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

OpenAI Google DeepMind Anthropic Cost frontier One model at each effort level
0 500 1,000 1,500 2,000 $0.01 $0.10 $1.00 $10 Cost per task, log scale GPT-6.2 Astra, medium Gemini 4 Hydrogen, medium Gemini 4 Argon, medium Claude Haiku 5.5, medium GPT-6 Luna, medium 0 500 1,000 1,500 2,000 $0.01 $0.10 $1.00 $10 Cost per task, log scale GPT-6.2 Astra, medium Gemini 4 Hydrogen, medium

Source: Official Crest Arcade Chess leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-6.2 Astra medium 1,758 $1.37
Gemini 4 Hydrogen medium 1,557 $1.18
GPT-6.1 Sol medium 1,508 $0.90
Gemini 4 Neon medium 1,497 $2.19
Gemini 4 Argon medium 1,489 $0.30
Gemini 4 Carbon medium 1,489 $0.87
GPT-6 Astra medium 1,429 $2.38
GPT-6 Sol medium 1,410 $0.88
Claude Opus 5.5 medium 1,346 $0.91
Claude Sonnet 5.5 medium 1,163 $0.47
Claude Haiku 5.5 medium 1,090 $0.05
GPT-6.1 Terra medium 1,085 $0.11
Claude Fable 5.1 medium 1,083 $2.90
gpt-oss-20b low 1,062
GPT-6 Luna medium 761 $0.02
Qwen3.8-Flash-Next low 502
Gemma 3 4B none 99
Qwen3.5-9B none 62
Qwen 3.8 27B none 62

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 Anthropic
Own leaderboard 7 leaders

Source: Official Crest Arcade Chess 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
0 500 1,000 1,500 2,000 2025 2026 2027 0 500 1,000 1,500 2,000 2025 2026 2027

Source: Official Crest Arcade Chess 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 (20 entries).
  • Pass Models scored 19 models we track, plus 1 other entry on its leaderboard.
  • Pass Newest score The newest is from November 2026, for Gemini 4 Neon.
  • Pass Frontier models 9 of the top 10 have a score, led by GPT-6.2 Astra (#1).
  • Note Headroom Scored as a rating, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on glimmerindex.com: 19 rows for models we track and 1 other entry 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