Benchmark Almanac

BoardBench 0.1

Claude Fable 5 91.4% leads since August 2026
Not in model cards
Version 0.1

Models play Yatzi, Scotland Yard, and Splendor. Game engines identify the best action at each decision and the scorer counts exact matches.

The maker scores ten distinct models across twelve settings. Claude Fable 5 matched the preferred action on 85 of 93 decisions.

Made by
BoardBench
Released
August 2026
Kind
Multi-agent games
Level
Everyday
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Actions
Interaction
Agentic
Environment
Yatzy, Scotland Yard, Splendor
Run by
Anyone
In views
Games, Agents
Strongest model tested
GPT-6 Astra #4 on the capability index
Last model added
September 2026
Scores refreshed
November 20th, 2026
10
Models scored on its own leaderboard

Claude Fable 5 leads with 91.4% among models released by September 2026

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.

The newest model with a score here came out in September 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Claude Fable 5 medium 91.4%
  2. 2 Claude Sonnet 5 medium 86.1%
  3. 3 GPT-6 Astra high 83.5%
  4. 4 Kimi K3 medium 82%
  5. 5 Grok 4.6 medium 82%
  6. 6 Claude Opus 5 medium 81.1%
  7. 7 Grok 4.5 medium 80.8%
  8. 8 DeepSeek V4 Flash 0731 high 80.7%
  9. 9 GPT-5.6 Luna high 77.7%
  10. 10 GPT-5.6 Sol medium 77.4%

Source: Official BoardBench 0.1 leaderboard

#ModelScoreSettingDateLinkSource
1 Claude Fable 5 91.4% medium Jun 2026 boardbench.dev Official leaderboard
2 Claude Sonnet 5 86.1% medium Jun 2026 boardbench.dev Official leaderboard
3 GPT-6 Astra 83.5% high Sep 2026 boardbench.dev Official leaderboard
4 Kimi K3 82% medium Jul 2026 boardbench.dev Official leaderboard
5 Grok 4.6 82% medium Aug 2026 boardbench.dev Official leaderboard
6 Claude Opus 5 81.1% medium Jul 2026 boardbench.dev Official leaderboard
7 Grok 4.5 80.8% medium Jul 2026 boardbench.dev Official leaderboard
8 DeepSeek V4 Flash 0731 80.7% high Jul 2026 boardbench.dev Official leaderboard
9 GPT-5.6 Luna 77.7% high Jul 2026 boardbench.dev Official leaderboard
10 GPT-5.6 Sol 77.4% medium Jul 2026 boardbench.dev Official leaderboard

The best score, 91.4%, costs $1.67 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.

Anthropic xAI DeepSeek Cost frontier One model at each effort level
0% 20% 40% 60% 80% 100% $0.01 $0.10 $1.00 $10 Cost per task, log scale Claude Fable 5, medium Claude Sonnet 5, medium Grok 4.6, medium 0% 20% 40% 60% 80% 100% $0.01 $0.10 $1.00 $10 Cost per task, log scale Claude Fable 5, medium Grok 4.6, medium

Source: Official BoardBench 0.1 leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5 medium 91.4% $1.67
Claude Sonnet 5 medium 86.1% $0.30
GPT-6 Astra high 83.5% $0.59
Kimi K3 medium 82% $0.14
Grok 4.6 medium 82% $0.07
Grok 4.6 high 81.8% $0.07
Claude Opus 5 medium 81.1% $1.66
Grok 4.5 medium 80.8% $0.06
DeepSeek V4 Flash 0731 high 80.7% $0.02
GPT-5.6 Luna high 77.7% $0.03
GPT-5.6 Sol medium 77.4% $0.33
Grok 4.5 low 77.3% $0.07

Number one over time

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

Anthropic
Own leaderboard Passed 85% in Aug 2026, by Claude Fable 5

Source: Official BoardBench 0.1 leaderboard

Claude Fable 5 on the official leaderboard passed 85% in June 2026

Models released before it came out already scored that high. 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% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Fable 5 0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Fable 5

Source: Official BoardBench 0.1 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.

Versions

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 9 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (10 entries).
  • Pass Models scored 10 models we track.
  • Pass Newest score The newest is from September 2026, for GPT-6 Astra.
  • Note Frontier models 1 of the top 10 has a score, which is fine for a retired test. The strongest tested is GPT-6 Astra (#4).
  • Note Headroom Saturated. The best is 91.4%, past 85% since June 2026.
Leaderboard import 3 pass
  • Pass Importer We read the board on boardbench.dev: 12 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.

How we check