KradleBench
Models play live multi-agent Minecraft games and act by writing code. The overall score weights collaboration, economy and conflict at 30% each, and logic and creativity at 5% each.
Kradle launched it on November 18, 2026 with 26 models and one skilled human player (battery kb-1.8.2). Runs last 2 to 15 minutes, and every model uses the same harness, prompt and skills.
- Made by
- Kradle
- Released
- November 2026
- Kind
- Multi-agent games
- Level
- Everyday
- Scored
- Settled by events
- Answers
- Agent task
- Measures
- Capability
- Language
- English
- Interaction
- Multi-agent
- Environment
- Minecraft
- Items from
- Live play or votes
- Refresh
- Live
- Run by
- A host
- In views
- Games, Agents, Kept fresh
- Strongest model tested
- GPT-6.2 Astra #1 on the capability index
- Last model added
- November 2026
- Scores refreshed
- November 20th, 2026
GPT-6.2 Astra leads with 78.3
Best score per model 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.
- 78.3 GPT-6.2 Astra
- 67.9 GPT-6.2 Sol
- 56.1 GPT-6 Astra
- 54.6 GPT-6.1 Sol
- 48.6 GPT-6.1 Terra
- 43.9
Gemini 4 Hydrogen
- 39.1 Claude Opus 5.5
- 37.8 Claude Fable 5.2
- 33.8 Claude Sonnet 5.5
- 28.6
Gemini 4 Argon
- 27.6
Muse Wonder
- 25.7
Muse Ember
- 24.3
Gemini 3.8 Flash
- 23.0 Grok 4.7
- 22.0
Gemini 4 Carbon
- 21.7
Muse Spark 1.3
- 21.2 GLM-5.3
- 19.8
DeepSeek V4.1 Flash
- 18.1 GPT-6 Luna
- 17.1
Gemma 4 31B IT
- 15.5 Kimi K3.1 Pro
- 15.2
Muse Spark Open
- 15.0 Claude Haiku 5.5
- 11.2
Nemotron 3.5 Super VL
- 11.0
Qwen4-Max
- 8.68
Qwen 3.8 Max
- 1 GPT-6.2 Astra 78.3
- 2 GPT-6.2 Sol 67.9
- 3 GPT-6 Astra 56.1
- 4 GPT-6.1 Sol 54.6
- 5 GPT-6.1 Terra 48.6
- 6
Gemini 4 Hydrogen 43.9
- 7 Claude Opus 5.5 39.1
- 8 Claude Fable 5.2 37.8
- 9 Claude Sonnet 5.5 33.8
- 10
Gemini 4 Argon 28.6
- 11
Muse Wonder 27.6
- 12
Muse Ember 25.7
- 13
Gemini 3.8 Flash 24.3
- 14 Grok 4.7 23.0
- 15
Gemini 4 Carbon 22.0
- 16
Muse Spark 1.3 21.7
- 17 GLM-5.3 21.2
- 18
DeepSeek V4.1 Flash 19.8
- 19 GPT-6 Luna 18.1
- 20
Gemma 4 31B IT 17.1
- 21 Kimi K3.1 Pro 15.5
- 22
Muse Spark Open 15.2
- 23 Claude Haiku 5.5 15.0
- 24
Nemotron 3.5 Super VL 11.2
- 25
Qwen4-Max 11.0
- 26
Qwen 3.8 Max 8.68
Source: Official KradleBench leaderboard
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-6.2 Astra | 78.3 | Nov 2026 | Official leaderboard | |
| 2 | GPT-6.2 Sol | 67.9 | Nov 2026 | Official leaderboard | |
| 3 | GPT-6 Astra | 56.1 | Sep 2026 | Official leaderboard | |
| 4 | GPT-6.1 Sol | 54.6 | Sep 2026 | Official leaderboard | |
| 5 | GPT-6.1 Terra | 48.6 | Oct 2026 | Official leaderboard | |
| 6 | | 43.9 | Nov 2026 | Official leaderboard | |
| 7 | Claude Opus 5.5 | 39.1 | Sep 2026 | Official leaderboard | |
| 8 | Claude Fable 5.2 | 37.8 | Nov 2026 | Official leaderboard | |
| 9 | Claude Sonnet 5.5 | 33.8 | Sep 2026 | Official leaderboard | |
| 10 | | 28.6 | Sep 2026 | Official leaderboard | |
| 11 | | 27.6 | Oct 2026 | Official leaderboard | |
| 12 | | 25.7 | Nov 2026 | Official leaderboard | |
| 13 | | 24.3 | Sep 2026 | Official leaderboard | |
| 14 | Grok 4.7 | 23.0 | Sep 2026 | Official leaderboard | |
| 15 | | 22.0 | Oct 2026 | Official leaderboard | |
| 16 | | 21.7 | Sep 2026 | Official leaderboard | |
| 17 | GLM-5.3 | 21.2 | Aug 2026 | Official leaderboard | |
| 18 | | 19.8 | Sep 2026 | Official leaderboard | |
| 19 | GPT-6 Luna | 18.1 | Sep 2026 | Official leaderboard | |
| 20 | | 17.1 | Apr 2026 | Official leaderboard | |
| 21 | Kimi K3.1 Pro | 15.5 | Oct 2026 | Official leaderboard | |
| 22 | | 15.2 | Oct 2026 | Official leaderboard | |
| 23 | Claude Haiku 5.5 | 15.0 | Oct 2026 | Official leaderboard | |
| 24 | | 11.2 | Oct 2026 | Official leaderboard | |
| 25 | | 11.0 | Nov 2026 | Official leaderboard | |
| 26 | | 8.68 | Aug 2026 | Official leaderboard |
Rankings by part
We read 5 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Collaboration, Economy and Conflict. GPT-6.2 Astra leads 4 of them, and GPT-6.2 Sol leads the other one. Arrows show how far a model moved from its overall rank among the models we track.
Collaboration
26 models we track have a score here.
- 1 GPT-6.2 Astra 74.3
- 2 GPT-6.2 Sol 70.2
- 3 GPT-6.1 Sol ↑1 61.7
- 4 GPT-6 Astra ↓1 59.7
- 5 Claude Fable 5.2 ↑3 56.0
- 6 GPT-6.1 Terra ↓1 54.9
- 7 Claude Opus 5.5 48.3
- 8
Gemini 4 Argon ↑2 47.6
- 9
Gemini 4 Hydrogen ↓3 43.8
- 10 Claude Sonnet 5.5 ↓1 40.4
Economy
26 models we track have a score here.
- 1 GPT-6.2 Astra 90.0
- 2 GPT-6.2 Sol 58.7
- 3 GPT-6 Astra 50.7
- 4 GPT-6.1 Sol 46.7
- 5
Gemini 4 Hydrogen ↑1 35.3
- 6 GPT-6.1 Terra ↓1 34.7
- 7
Gemma 4 31B IT ↑13 25.3
- 8
DeepSeek V4.1 Flash ↑10 22.7
- 9 Claude Sonnet 5.5 20.7
- 10
Muse Wonder ↑1 17.3
Conflict
26 models we track have a score here.
- 1 GPT-6.2 Sol ↑1 65.0
- 2 GPT-6.2 Astra ↓1 63.3
- 3 GPT-6 Astra 52.7
- 4
Gemini 4 Hydrogen ↑2 52.5
- 5 GPT-6.1 Sol ↓1 52.2
- 6 GPT-6.1 Terra ↓1 52.0
- 7 Claude Opus 5.5 47.3
- 8 Claude Sonnet 5.5 ↑1 34.7
- 9 Grok 4.7 ↑5 33.8
- 10 Claude Fable 5.2 ↓2 32.7
Logic
26 models we track have a score here.
- 1 GPT-6.2 Astra 100.0
- 2 GPT-6.2 Sol 97.7
- 3 Claude Fable 5.2 ↑5 67.3
- 4 GPT-6 Astra ↓1 46.3
- 5 Claude Opus 5.5 ↑2 34.7
- 6
Gemini 4 Argon ↑4 34.7
- 7 Claude Sonnet 5.5 ↑2 32.3
- 8 GPT-6.1 Sol ↓4 32.3
- 9
Gemini 4 Carbon ↑6 32.3
- 10 GPT-6.1 Terra ↓5 30.0
Creativity
26 models we track have a score here.
- 1 GPT-6.2 Astra 100.0
- 2 GPT-6 Astra ↑1 98.5
- 3 GPT-6.2 Sol ↓1 98.1
- 4 GPT-6.1 Sol 97.4
- 5
Gemini 4 Carbon ↑10 95.5
- 6 GPT-6.1 Terra ↓1 93.4
- 7
Gemini 4 Argon ↑3 88.0
- 8 Claude Fable 5.2 84.4
- 9 Claude Opus 5.5 ↓2 78.3
- 10 Kimi K3.1 Pro ↑11 71.7
The best score, 78.3, costs $1.26 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.
Source: Official KradleBench leaderboard
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| GPT-6.2 Astra | 78.3 | $1.26 | |||
| GPT-6.2 Sol | 67.9 | $0.35 | |||
| GPT-6 Astra | 56.1 | $1.55 | |||
| GPT-6.1 Sol | 54.6 | $0.32 | |||
| GPT-6.1 Terra | 48.6 | $0.16 | |||
| 43.9 | $0.53 | ||||
| Claude Opus 5.5 | 39.1 | $0.95 | |||
| Claude Fable 5.2 | 37.8 | $1.35 | |||
| Claude Sonnet 5.5 | 33.8 | $0.43 | |||
| 28.6 | $0.41 | ||||
| 27.6 | $1.13 | ||||
| 25.7 | $0.06 | ||||
| 24.3 | $0.27 | ||||
| Grok 4.7 | 23.0 | $0.63 | |||
| 22.0 | $0.30 | ||||
| 21.7 | $0.42 | ||||
| GLM-5.3 | 21.2 | $0.10 | |||
| 19.8 | $0.05 | ||||
| GPT-6 Luna | 18.1 | $0.02 | |||
| 17.1 | $0.08 | ||||
| Kimi K3.1 Pro | 15.5 | $0.47 | |||
| 15.2 | $0.42 | ||||
| Claude Haiku 5.5 | 15.0 | $0.03 | |||
| 11.2 | $0.11 | ||||
| 11.0 | $0.16 | ||||
| 8.68 | $0.13 |
2 models have held number one on the official leaderboard since its release in November 2026
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official KradleBench leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official KradleBench 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
- 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
9 of 10 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 4 pass 1 note
- Pass Independent runs Scores from the official leaderboard (26 entries).
- Pass Models scored 26 models we track.
- Pass Newest score The newest is from November 2026, for GPT-6.2 Astra.
- Pass Frontier models 9 of the top 10 have a score, led by GPT-6.2 Astra (#1).
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on kradle.ai: 26 rows for models we track 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.