Back to Benchmarks

DeepSeek R1 Distill Llama 70B vs GPT-4.1 nano

DeepSeek R1 Distill Llama 70B

DeepSeekvia DeepInfra

vs

GPT-4.1 nano

OpenAI

DeepSeek R1 Distill Llama 70B(DeepInfra) wins 9 of 11

Index Scores

16.0
+3.0
Intelligence
13.0
11.4
+0.2
Coding
11.2
53.7
+29.7
Math
24.0

Benchmarks

79.5
+13.8
MMLU Pro
65.7
40.2
GPQA
+11.0
51.2
6.1
+2.2
HLE
3.9
26.6
LiveCode
+6.0
32.6
31.2
+5.3
SciCode
25.9
93.5
+8.7
MATH-500
84.8
67.0
+43.3
AIME
23.7

Speed

42
+7
TPS
35
Shared
Winning delta

crafted by bart stefanski