Back to Benchmarks

Claude 4.5 Haiku (Reasoning) vs gpt-oss-120B (high)

Claude 4.5 Haiku (Reasoning)

Anthropic

vs

gpt-oss-120B (high)

OpenAIvia Groq

gpt-oss-120B (high)(Groq) wins 6 of 11

Index Scores

37.1
+3.8
Intelligence
33.3
32.6
+4.0
Coding
28.6
83.7
Math
+9.7
93.4

Benchmarks

76.0
MMLU Pro
+4.8
80.8
67.2
GPQA
+11.0
78.2
9.7
HLE
+8.8
18.5
61.5
LiveCode
+26.3
87.8
43.3
+4.4
SciCode
38.9
0.0
MATH-500
0.0
0.0
AIME
0.0

Speed

66
TPS
+283
349
Shared
Winning delta

crafted by bart stefanski