BenchLeader
Menu
Best LLMsChoose a modelCoding plansLeaderboardsModelsBenchmarksProvidersReleases7

ForecastBench leaderboard

As of 6 Nov 2026, the top score on ForecastBench is 67.0% by GPT-5.5 (OpenAI), ahead of Claude Opus 5 at 66.5%, across 85 model configurations with a published result (ForecastBench, result dated 23 Apr 2026). Models forecast real-world questions; scores accrue as the questions resolve, compared with superforecasters. ForecastBench is shown for reference only and does not count toward the BenchLeader Index.

Key facts (as of 6 Nov 2026)

  • GPT-5.5 holds the top ForecastBench score at 67.0% (ForecastBench, 23 Apr 2026).
  • Claude Opus 5 is second on ForecastBench at 66.5% (ForecastBench, 24 Jul 2026).
  • ForecastBench: Forecasting real future events, scored as they resolve (ForecastBench).
  • 85 model configurations have a published ForecastBench result on BenchLeader (BenchLeader, 6 Nov 2026).
  • ForecastBench does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).

Forecasting real future events, scored as they resolve.

Published by
ForecastBenchdata via Epoch AI Benchmarking Hub
Category
Reasoning
Index weight
Reference only
Models
85
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

Models forecast real-world questions; scores accrue as the questions resolve, compared with superforecasters.

How it is scored

Overall Brier-based score, as published by ForecastBench.

What to keep in mind

Scores change as events resolve; a lower Brier score is better and the site rescales it.

Top 15
  1. 1GPT-5.5 (high)67.0%
  2. 2Claude Opus 5 (max)66.5%
  3. 3GPT-5.6 Sol (medium)66.5%
  4. 4Claude Opus 4.8 (max)66.4%
  5. 5Gemini 3.7 Flash (high)66.2%
  6. 6Claude Fable 5 (max)66.0%
  7. 7Grok 4.5 (high)66.0%
  8. 8Gemini 3.6 Flash (medium)65.9%
  9. 9Claude Fable 5 (high)65.7%
  10. 10GPT-5.6 Sol (pro)65.6%
  11. 11GPT-5.6 Sol (max)65.5%
  12. 12Gemini 3.1 Pro (high)65.5%
  13. 13Grok 4.6 (xhigh)65.5%
  14. 14GPT-565.3%
  15. 15Gemini 3 Pro65.2%
85 of 85
#
1GPT-5.5high67.0%63.72026-04-23
2Claude Opus 5max66.5%65.42026-07-24
3GPT-5.6 Solmedium66.5%61.52026-07-09
4Claude Opus 4.8max66.4%60.22026-05-28
5Gemini 3.7 Flashhigh66.2%62.22026-08-13
6Claude Fable 5max66.0%65.32026-06-09
7Grok 4.5high66.0%56.92026-07-08
8Gemini 3.6 Flashmedium65.9%–2026-07-21
9Claude Fable 5high65.7%64.32026-06-09
10GPT-5.6 Solpro mode65.6%–2026-07-09
11GPT-5.6 Solmax65.5%65.42026-07-09
12Gemini 3.1 Prohigh65.5%60.12026-02-19
13Grok 4.6xhigh65.5%57.92026-08-12
14GPT-565.3%57.72025-08-07
15Gemini 3 Pro65.2%60.22025-11-18
16Grok 465.0%58.32025-07-09
17Claude Opus 4.564.9%57.92025-11-24
18Grok 4 Fast64.9%–2025-09-19
19Grok 4.1 Fastthinking64.8%55.62025-11-19
20GPT-4.564.8%50.82025-02-27
21Kimi K3max64.7%62.12026-07-16
22Gemini 3 Flash64.7%56.92025-12-17
23Claude Opus 4.164.7%51.42025-08-05
24o364.6%61.02025-04-16
25GPT-5.564.5%61.72026-04-23
26Grok 4.20thinking64.5%58.42026-02-17
27Gemini 2.5 Pro64.4%54.52025-03-31
28Gemini 3.5 Flash64.4%48.62026-05-19
29Gemini 3.1 Pro64.3%60.52026-02-19
30Claude Sonnet 4.664.3%58.22026-02-17
31MiniMax M364.3%56.42026-06-01
32GPT-5.264.2%57.92025-12-11
33o4-mini64.2%56.92025-04-16
34Claude Sonnet 464.2%50.52025-05-22
35Claude Sonnet 4.564.1%53.72025-09-29
36Qwen3 235B A22B64.1%50.02025-04-29
37Claude 3.7 Sonnet64.0%49.92025-02-24
38o3-mini63.8%50.72025-01-31
39GLM-563.7%54.82026-02-11
40Claude Sonnet 563.5%55.92026-06-30
41Claude Opus 4.663.4%62.32026-02-05
42Grok 4.363.4%54.52026-04-17
43Claude Haiku 4.563.4%47.62025-10-15
44GPT-5 mini63.3%56.42025-08-07
45Grok 4.2063.3%–2026-02-17
46Claude 3.5 Sonnet63.2%47.12024-10-22
47GPT-4 Turbo63.2%41.22024-04-09
48Claude Opus 4.763.1%63.12026-04-16
49GPT-5.463.1%58.02026-03-05
50Gemini 2.5 Flash62.9%52.72025-06-17
51Gemini 3.1 Flash Lite62.8%52.42026-03-03
52Claude Opus 4.862.7%60.02026-05-28
53Kimi K2 090562.7%51.82025-09-05
54DeepSeek V362.6%44.42024-12-26
55Llama 3.1 405B62.6%41.92024-07-23
56Claude 3 Opus62.4%39.02024-02-29
57GPT-5.162.3%56.82025-11-13
58GPT-4.162.3%48.52025-04-14
59Qwen3 235B A22B 2507thinking62.2%51.92025-07-25
60Kimi K262.2%51.02025-07-12

More reasoning benchmarks · All LLM benchmarks

Related benchmarks

Top head-to-heads on ForecastBench

Cite as: BenchLeader, “ForecastBench leaderboard”, data as of 6 Nov 2026, https://www.benchleader.com/benchmarks/forecastbench.

ForecastBench: questions

What does ForecastBench measure?
Models forecast real-world questions; scores accrue as the questions resolve, compared with superforecasters. Scores are reported in percent of tasks solved; higher is better.
What is the top score on ForecastBench?
As of 6 Nov 2026, the highest score on ForecastBench is 67.0% by GPT-5.5 (result dated 23 Apr 2026), ahead of Claude Opus 5 at 66.5%. Scores come from ForecastBench and are re-read daily.
Which AI model leads ForecastBench?
GPT-5.5 leads ForecastBench with 67.0% as of 6 Nov 2026 (result dated 23 Apr 2026), ahead of Claude Opus 5 at 66.5%.
How many models have ForecastBench results?
85 model configurations have a ForecastBench result on BenchLeader, all taken from ForecastBench via Epoch AI Benchmarking Hub.
Who runs ForecastBench and how often is it updated?
ForecastBench is published by ForecastBench. BenchLeader re-reads the published results once a day and records the date each result was published.
Does ForecastBench count toward the BenchLeader Index?
No. ForecastBench is shown for reference but left out of the composite index.