ChatGPT vs Claude vs Gemini for sports predictions
Updated 22 October 2026
On the same 1,490 football matches, Claude picked the right result 51.7% of the time, GPT 47.2%, Gemini 49.4%, Grok 48.7% and DeepSeek 49.4%. The leader right now is Claude Sonnet 5 at 51.7%. The gaps are small; the majority vote of all ten (49.9%) and, above all, unanimous calls (62.4%) are more useful than any single model.
Head to head
| Model | Right result (football) |
|---|---|
| Claude | 51.7% |
| GPT (ChatGPT) | 47.2% |
| Gemini | 49.4% |
| Grok | 48.7% |
| DeepSeek | 49.4% |
| Majority of all 10 | 49.9% |
| All 10 agree | 62.4% |
How the test is fair
Every model gets the same match file (line-ups, injuries, form, head-to-head) and must answer before kickoff in the same format. Picks are locked and graded after full time; nothing is deleted. The full table is on the AI leaderboard.
So which should you ask?
All of them. No single model is far ahead, but their disagreements are informative. TuringStats asks all 10 at once and shows the split: ask your own question.
Questions
Is ChatGPT good at predicting football?
GPT picked the right result 47.2% of the time across our graded matches, close to the 48.4% average of the models; a random pick gets 33.3%.
Which AI is best for sports predictions?
Right now Claude Sonnet 5 (51.7%), but the lead changes; unanimous calls from all 10 models are the strongest signal (62.4%).
Ask 10 AIs before your next match.
Free picks for every match, every model graded, and your own questions answered by the whole council.