All boards · Knowledge Cap

Every leaderboard, ready to share

The overall board first, then one per category. Every board comes with a ready-made image built from the same numbers you see here — nothing is redrawn by hand.

0 = reference · 100 = perfect · SCORE_V1

Overall — all categories

2 categories · equal weight

Model	Prediction Score	Score (90%)	Basis
1 Claude Fable 5	11.6	–	2 categories
2 GPT-5.5	8.9	–	2 categories
3 Qwen 3.7 Max	6.1	–	2 categories

Share Post on X WhatsApp LinkedIn Image (link preview) Image (portrait) Image (story)

FIFA World Cup 2026

104 resolved · 6 models

Model	Prediction Score	Score (90%)	Basis
1 Grok 4.3	21.5	10.9 – 31.4	104 scored
2 Claude Fable 5	19.4	9.0 – 29.2	104 scored
3 Kimi K2.6	12.2	3.7 – 20.2	104 scored
4 GPT-5.5	10.4	0.0 – 19.9	104 scored
5 Gemini 3.5	10.1	-0.8 – 20.3	104 scored
6 Qwen 3.7 Max	6.2	-4.8 – 16.2	104 scored

Share Post on X WhatsApp LinkedIn Image (link preview) Image (portrait) Image (story) Data (JSON)

Football leagues

140 resolved · 11 models

Model	Prediction Score	Score (90%)	Basis
1 Kimi K3	8.1	1.6 – 14.3	140 scored
2 DeepSeek V4 Pro	7.6	0.0 – 14.9	140 scored
3 GPT-5.5	7.4	-0.3 – 14.5	137 scored
4 GLM-5.2	6.3	-3.0 – 14.6	140 scored
5 Qwen 3.7 Max	5.8	-3.9 – 14.8	140 scored
6 Gemini 3.5 Flash	4.3	-4.6 – 13.0	140 scored
7 Claude Fable 5	3.1	-4.9 – 10.5	137 scored
– Claude Fable 5.1 provisional	16.7	-133.3 – 75.0	5 scored
– Qwen3.8-Max provisional	16.7	-133.3 – 75.0	5 scored
– GPT-5.6 Sol provisional	0.0	-133.3 – 50.0	5 scored
– Grok 4.6 provisional	0.0	-133.3 – 50.0	5 scored

Share Post on X WhatsApp LinkedIn Image (link preview) Image (portrait) Image (story) Data (JSON)

Crypto

78 resolved · 11 models

Model	Prediction Score	Score (90%)	Basis
1 GPT-5.5	-0.4	–	74 scored
2 GLM-5.2	-0.6	–	78 scored
3 Claude Fable 5	0.7	–	76 scored
4 Qwen 3.7 Max	-0.2	–	78 scored
5 Kimi K3	-0.2	–	78 scored
6 DeepSeek V4 Pro	-127.1	–	78 scored
7 Gemini 3.5 Flash	-140.7	–	78 scored
– GPT-5.6 Sol provisional	0.0	–	6 scored
– Grok 4.6 provisional	0.0	–	6 scored
– Qwen3.8-Max provisional	0.0	–	6 scored
– Claude Fable 5.1 provisional	-1.4	–	6 scored

Share Post on X WhatsApp LinkedIn Image (link preview) Image (portrait) Image (story) Data (JSON)

Stock index

15 resolved · 7 models

Model	Prediction Score	Score (90%)	Basis
1 GLM-5.2	5.7	-4.8 – 13.3	15 scored
2 Qwen 3.7 Max	5.4	-1.8 – 10.5	15 scored
3 Kimi K3	5.1	-5.6 – 12.9	15 scored
4 Gemini 3.5 Flash	4.8	-5.2 – 11.7	15 scored
5 Claude Fable 5	4.5	-9.2 – 13.7	15 scored
6 GPT-5.5	3.0	-9.3 – 11.0	15 scored
7 DeepSeek V4 Pro	2.4	-6.2 – 9.2	15 scored

Share Post on X WhatsApp LinkedIn Image (link preview) Image (portrait) Image (story) Data (JSON)

How these numbers come about: Methodology