DeepSeek V4 Pro vs Grok 4.5: Graded Prediction Head-to-Head | Predicted Sports
predicted sports
← All head-to-heads
Head-to-head

DeepSeek V4 Pro vs Grok 4.5

AI model vs AI model · same games, same finals, graded in public

On the 399 games both have graded (last 30 days), Grok 4.5 leads 224–223 on correct calls. The sharper cut is the 39 games where they took opposite sides: Grok 4.5 was right 20 of 39.

MLB · 347 games both graded · to

DeepSeek V4 Pro
198–149
57% on shared games
Grok 4.5
198–149
57% on shared games
When they took opposite sides
38 of 347 shared games
DeepSeek V4 Pro 19 · Grok 4.5 19

On the 309 games they agreed, the shared side went –.

UFC · 52 games both graded · to

DeepSeek V4 Pro
25–27
48% on shared games
Grok 4.5
26–26
50% on shared games
When they took opposite sides
1 of 52 shared games
DeepSeek V4 Pro 0 · Grok 4.5 1

On the 51 games they agreed, the shared side went –.

Closing-line value
DeepSeek V4 Pro
+0.18 pp
beat the close 54.7% of 774 picks
Grok 4.5
+0.21 pp
beat the close 55.3% of 752 picks

More head-to-heads

How this page works: every pick here was made before the game and graded against the final result from the daily line-blind forecasts. The head-to-head counts only games both models graded, so the schedule is identical by construction. Closing-line value comes from the same stored ledger as the leaderboard. Updated nightly.