This website uses cookies for authentication, support chat, and analytics. Only authentication cookies are turned on by default.
Animate this data
Bar chart race of phi-4, gpt-6-astra_max, gemini-3.8-flash_high and 7 more, by GPQA Diamond.
Create it free
Takes a free account
phi-4
BALROG
0.12 %
Chess Puzzles
0.01 %
GPQA Diamond
0.56 %
Lech Mazur Writing
0.06 %
Live bench
0.42 %
Math Level 5
0.65 %
Mmlu
0.85 %
OTIS Mock AIME 2024 2025
0.14 %
Training Compute
932,020,150,000.00T FLOP
phi-4 benchmark results.