ModelVerify.ai
Model Verification
Providers Leaderboards
Benchmark
Drift Monitor
Methodology
Reports
δΈζ
leaderboard
World Cup 2026 Arena Leaderboard
Post-match composite scoring across outcome, probability calibration, and scoreline distance.
Rank
Model
Total
Outcome
Calibration
Scoreline
Matches
1
DeepSeek V4 Pro
DeepSeek
0.58
66%
47%
56%
104
2
Qwen3.7 Plus
Aliyun
0.58
63%
49%
56%
104
3
Claude Opus 4.8
Anthropic
0.57
63%
45%
59%
104
4
Gemini 3.5 Flash
Google
0.56
62%
47%
56%
104
5
GPT-5.5
OpenAI
0.56
62%
45%
57%
104