Dutch Book Bench
Models assign probabilities to related events in separate conversations. The grader checks whether inconsistent probabilities let a bettor guarantee a profit.
Kaggle publishes scores for 19 distinct model versions.
- Made by
- Yi Yu
- Released
- October 2026
- Kind
- Real-world events
- Level
- University
- Scored
- Checked automatically
- Answers
- Prediction
- Tasks
- All public
- Measures
- Capability
- Subject
- Mathematics
- Language
- English
- Interaction
- Single turn
- Refresh
- One fixed set
- Run by
- Anyone
- Strongest model tested
- Claude Sonnet 5.5 #9 on the capability index
- Last model added
- October 2026
- Scores refreshed
- November 20th, 2026
Gemini 3.8 Flash leads with 89.2% Gemini 3.8 Flash leads with 89.2%
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.
- 89.2
Gemini 3.8 Flash
- 83.8 GPT-5.5
- 75.7
Gemini 3.6 Flash
- 75.7 Claude Sonnet 5.5
- 73.0 GPT-5.6 Luna
- 70.3
Gemini 3.1 Flash-Lite
- 70.3
Gemini 3.5 Flash-Lite
- 70.3 Claude Haiku 5.5
- 67.6
Gemini 3 Flash
- 67.6
Gemini 3.1 Pro
- 67.6 Claude Sonnet 5
- 66.7
Gemma 4 31B IT
- 64.9 Claude Haiku 4.5
- 64.9 GPT-5.4 Mini
- 62.2 Qwen 3 235B A22B Instruct 2506
- 59.5 Qwen 3 Next 80B Instruct
- 56.8 Grok 4.20
- 54.0 GPT-5.4 Nano
- 51.4
Qwen3-Coder-480B-A35B
- 1
Gemini 3.8 Flash 89.2%
- 2 GPT-5.5 83.8%
- 3
Gemini 3.6 Flash 75.7%
- 4 Claude Sonnet 5.5 75.7%
- 5 GPT-5.6 Luna 73%
- 6
Gemini 3.1 Flash-Lite 70.3%
- 7
Gemini 3.5 Flash-Lite 70.3%
- 8 Claude Haiku 5.5 70.3%
- 9
Gemini 3 Flash 67.6%
- 10
Gemini 3.1 Pro 67.6%
- 11 Claude Sonnet 5 67.6%
- 12
Gemma 4 31B IT 66.7%
- 13 Claude Haiku 4.5 64.9%
- 14 GPT-5.4 Mini 64.9%
- 15 Qwen 3 235B A22B Instruct 2506 62.2%
- 16 Qwen 3 Next 80B Instruct 59.5%
- 17 Grok 4.20 non-reasoning 56.8%
- 18 GPT-5.4 Nano 54%
- 19
Qwen3-Coder-480B-A35B 51.4%
Source: Official Dutch Book Bench leaderboard
- 89.2
Gemini 3.8 Flash
- 83.8 GPT-5.5
- 75.7
Gemini 3.6 Flash
- 75.7 Claude Sonnet 5.5
- 73.0 GPT-5.6 Luna
- 70.3
Gemini 3.1 Flash-Lite
- 70.3
Gemini 3.5 Flash-Lite
- 70.3 Claude Haiku 5.5
- 67.6
Gemini 3 Flash
- 67.6
Gemini 3.1 Pro
- 67.6 Claude Sonnet 5
- 66.7
Gemma 4 31B IT
- 64.9 Claude Haiku 4.5
- 64.9 GPT-5.4 Mini
- 56.8 Grok 4.20
- 54.0 GPT-5.4 Nano
- 51.4
Qwen3-Coder-480B-A35B
- 1
Gemini 3.8 Flash 89.2%
- 2 GPT-5.5 83.8%
- 3
Gemini 3.6 Flash 75.7%
- 4 Claude Sonnet 5.5 75.7%
- 5 GPT-5.6 Luna 73%
- 6
Gemini 3.1 Flash-Lite 70.3%
- 7
Gemini 3.5 Flash-Lite 70.3%
- 8 Claude Haiku 5.5 70.3%
- 9
Gemini 3 Flash 67.6%
- 10
Gemini 3.1 Pro 67.6%
- 11 Claude Sonnet 5 67.6%
- 12
Gemma 4 31B IT 66.7%
- 13 Claude Haiku 4.5 64.9%
- 14 GPT-5.4 Mini 64.9%
- 15 Grok 4.20 non-reasoning 56.8%
- 16 GPT-5.4 Nano 54%
- 17
Qwen3-Coder-480B-A35B 51.4%
Source: Official Dutch Book Bench leaderboard
| # | Entry | Score | Setting | Released | Run on | Source |
|---|---|---|---|---|---|---|
| 1 | | 89.2% | Sep 2026 | Oct 2026 | Official leaderboard | |
| 2 | GPT-5.5 | 83.8% | Apr 2026 | Oct 2026 | Official leaderboard | |
| 3 | | 75.7% | Jul 2026 | Oct 2026 | Official leaderboard | |
| 4 | Claude Sonnet 5.5 | 75.7% | Sep 2026 | Oct 2026 | Official leaderboard | |
| 5 | GPT-5.6 Luna | 73% | Jul 2026 | Oct 2026 | Official leaderboard | |
| 6 | | 70.3% | Mar 2026 | Oct 2026 | Official leaderboard | |
| 7 | | 70.3% | Jul 2026 | Oct 2026 | Official leaderboard | |
| 8 | Claude Haiku 5.5 | 70.3% | Oct 2026 | Oct 2026 | Official leaderboard | |
| 9 | | 67.6% | Dec 2025 | Oct 2026 | Official leaderboard | |
| 10 | | 67.6% | Feb 2026 | Oct 2026 | Official leaderboard | |
| 11 | Claude Sonnet 5 | 67.6% | Jun 2026 | Oct 2026 | Official leaderboard | |
| 12 | | 66.7% | Apr 2026 | Oct 2026 | Official leaderboard | |
| 13 | Claude Haiku 4.5 | 64.9% | Oct 2025 | Oct 2026 | Official leaderboard | |
| 14 | GPT-5.4 Mini | 64.9% | Mar 2026 | Oct 2026 | Official leaderboard | |
| 15 | Qwen 3 235B A22B Instruct 2506 | 62.2% | Oct 2026 | Official leaderboard | ||
| 16 | Qwen 3 Next 80B Instruct | 59.5% | Oct 2026 | Official leaderboard | ||
| 17 | Grok 4.20 | 56.8% | non-reasoning | Feb 2026 | Oct 2026 | Official leaderboard |
| 18 | GPT-5.4 Nano | 54% | Mar 2026 | Oct 2026 | Official leaderboard | |
| 19 | | 51.4% | Jul 2025 | Oct 2026 | Official leaderboard |
| # | Model | Score | Setting | Released | Run on | Source |
|---|---|---|---|---|---|---|
| 1 | | 89.2% | Sep 2026 | Oct 2026 | Official leaderboard | |
| 2 | GPT-5.5 | 83.8% | Apr 2026 | Oct 2026 | Official leaderboard | |
| 3 | | 75.7% | Jul 2026 | Oct 2026 | Official leaderboard | |
| 4 | Claude Sonnet 5.5 | 75.7% | Sep 2026 | Oct 2026 | Official leaderboard | |
| 5 | GPT-5.6 Luna | 73% | Jul 2026 | Oct 2026 | Official leaderboard | |
| 6 | | 70.3% | Mar 2026 | Oct 2026 | Official leaderboard | |
| 7 | | 70.3% | Jul 2026 | Oct 2026 | Official leaderboard | |
| 8 | Claude Haiku 5.5 | 70.3% | Oct 2026 | Oct 2026 | Official leaderboard | |
| 9 | | 67.6% | Dec 2025 | Oct 2026 | Official leaderboard | |
| 10 | | 67.6% | Feb 2026 | Oct 2026 | Official leaderboard | |
| 11 | Claude Sonnet 5 | 67.6% | Jun 2026 | Oct 2026 | Official leaderboard | |
| 12 | | 66.7% | Apr 2026 | Oct 2026 | Official leaderboard | |
| 13 | Claude Haiku 4.5 | 64.9% | Oct 2025 | Oct 2026 | Official leaderboard | |
| 14 | GPT-5.4 Mini | 64.9% | Mar 2026 | Oct 2026 | Official leaderboard | |
| 15 | Grok 4.20 | 56.8% | non-reasoning | Feb 2026 | Oct 2026 | Official leaderboard |
| 16 | GPT-5.4 Nano | 54% | Mar 2026 | Oct 2026 | Official leaderboard | |
| 17 | | 51.4% | Jul 2025 | Oct 2026 | Official leaderboard |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official Dutch Book Bench leaderboard
Gemini 3.8 Flash on the official leaderboard passed 85% in September 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.
Source: Official Dutch Book Bench 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 Forecasting
- ForecastBench In use Models give probabilities for 500 auto-generated questions per round about future events, drawn from prediction markets and time-series datasets. Forecasts are scored with difficulty-adjusted Brier scores once the questions resolve. GPT-5.5 0.11
- Prophet Arena In use Models and agents forecast open Kalshi prediction-market events and are scored once the events resolve. The live board ranks a difficulty-adjusted Arena score and also reports calibration error and market return. Claude Opus 5.5 86.2
- Guesswork (The Aggregate) Not in model cards Models predict benchmark scores before publication. The maker ranks forecasts by mean absolute error after scaling each error by the variation in that benchmark’s scores. GPT-6 Astra 0.56
- AGI-Pilled Bench Not in model cards Models forecast when AI will reach milestones that already happened in 2026. The score is the median years after 2026 across the forecast tasks. Claude Fable 5.2 6.50
- Echo General AI Prediction Leaderboard Not in model cards Models forecast unresolved events drawn from prediction markets, synthesized news questions and expert-written questions. Forecasts made at aligned times are compared in pairwise battles, and an Elo fit ranks the models.
DeepSeek-V4-Pro 1,018
- SynthBench Not in model cards Models answer survey questions. The index compares their answer distributions with human responses across four datasets, using random answers and human agreement as reference points.
Qwen3.8 Flash 32.2
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 (19 entries).
- Pass Models scored 17 models we track, plus 2 other entries on its leaderboard.
- Pass Newest score The newest is from October 2026, for Claude Haiku 5.5.
- Note Frontier models 1 of the top 10 has a score, which is fine for a retired test. The strongest tested is Claude Sonnet 5.5 (#9).
- Note Headroom Saturated. The best is 89.2%, past 85% since September 2026.
Leaderboard import 3 pass
- Pass Importer We read the board on kaggle.com: 17 rows for models we track and 2 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.