BTF-3 leaderboard
As of 6 Nov 2026, the top score on BTF-3 is 14.9%, shared by GLM 5.3 Flash and GLM 5.3, ahead of GPT-5.5 at 14.3%, across 13 model configurations with a published result (Epoch AI Benchmarking Hub, result dated 20 Aug 2026). Binary and numeric forecasting questions scored with Brier and ranked probability scores, pooled. BTF-3 is shown for reference only and does not count toward the BenchLeader Index.
Key facts (as of 6 Nov 2026)
- GLM 5.3 Flash and GLM 5.3 share the top BTF-3 score at 14.9% (Epoch AI Benchmarking Hub, 6 Nov 2026).
- BTF-3: Forecasting tournament pooled score (Epoch AI Benchmarking Hub).
- 13 model configurations have a published BTF-3 result on BenchLeader (BenchLeader, 6 Nov 2026).
- BTF-3 does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).
Forecasting tournament pooled score.
- Published by
- Epoch AI Benchmarking Hub
- Category
- Reasoning
- Index weight
- Reference only
- Models
- 13
- Data as of
- 6 Nov 2026
CC BY 4.0 — Epoch AI, ‘AI Benchmarking Hub’, epoch.ai/benchmarks. Mirrored boards credit their original publishers.
What the test looks like
Binary and numeric forecasting questions scored with Brier and ranked probability scores, pooled.
How it is scored
Pooled score, as published.
What to keep in mind
Very few models; scores depend on question resolution.
- 1GLM 5.3 Flash14.9%
- 2GLM 5.314.9%
- 3GPT-5.5 (high)14.3%
- 4Claude Sonnet 5 (xhigh)14.2%
- 5Muse Spark 1.3 (xhigh)14.1%
- 6Claude Opus 4.8 (high)14.0%
- 7GPT-5.6 Sol (max)13.7%
- 8GPT-5.6 Sol (medium)13.7%
- 9GPT-5.6 Sol (high)13.5%
- 10GPT-6 Astra (low)13.5%
- 11Claude Opus 4.8 (xhigh)13.2%
- 12Claude Fable 5 (high)13.0%
- 13Claude Opus 5 (xhigh)12.0%
| # | ||||
|---|---|---|---|---|
| 1 |  GLM 5.3 Flashopen ↗ | 14.9% | 56.8 | 2026-08-20 |
| 2 | GLM 5.3open ↗ | 14.9% | – | 2026-08-14 |
| 3 |  GPT-5.5high | 14.3% | 63.7 | 2026-04-23 |
| 4 |  Claude Sonnet 5xhigh | 14.2% | 55.6 | 2026-06-30 |
| 5 |  Muse Spark 1.3xhigh | 14.1% | 58.5 | 2026-09-02 |
| 6 | Claude Opus 4.8high | 14.0% | 60.6 | 2026-05-28 |
| 7 | GPT-5.6 Solmax | 13.7% | 65.4 | 2026-07-09 |
| 8 | GPT-5.6 Solmedium | 13.7% | 61.5 | 2026-07-09 |
| 9 | GPT-5.6 Solhigh | 13.5% | 63.6 | 2026-07-09 |
| 10 | GPT-6 Astralow | 13.5% | 62.7 | 2026-09-03 |
| 11 | Claude Opus 4.8xhigh | 13.2% | – | 2026-05-28 |
| 12 | Claude Fable 5high | 13.0% | 64.3 | 2026-06-09 |
| 13 | Claude Opus 5xhigh | 12.0% | 60.2 | 2026-07-24 |
More reasoning benchmarks · All LLM benchmarks
Related benchmarks
Top head-to-heads on BTF-3
Cite as: BenchLeader, “BTF-3 leaderboard”, data as of 6 Nov 2026, https://www.benchleader.com/benchmarks/btf3.
BTF-3: questions
- What does BTF-3 measure?
- Binary and numeric forecasting questions scored with Brier and ranked probability scores, pooled. Scores are reported in percent of tasks solved; higher is better.
- What is the top score on BTF-3?
- As of 6 Nov 2026, the highest score on BTF-3 is 14.9%, shared by GLM 5.3 Flash and GLM 5.3, ahead of GPT-5.5 at 14.3%. Scores come from Epoch AI Benchmarking Hub and are re-read daily.
- Which AI model leads BTF-3?
- GLM 5.3 Flash and GLM 5.3 tie for first on BTF-3 at 14.9% as of 6 Nov 2026, ahead of GPT-5.5 at 14.3%.
- How many models have BTF-3 results?
- 13 model configurations have a BTF-3 result on BenchLeader, all taken from Epoch AI Benchmarking Hub.
- Who runs BTF-3 and how often is it updated?
- BTF-3 is published by Epoch AI Benchmarking Hub. BenchLeader re-reads the published results once a day and records the date each result was published.
- Does BTF-3 count toward the BenchLeader Index?
- No. BTF-3 is shown for reference but left out of the composite index.