Side-by-side benchmark comparison across knowledge, coding, math, and reasoning.
Claude 3 Opus wins overall with a score of 51 vs 48 (3 point difference).Claude 3 Opus wins 4 out of 4 categories.
Claude 3 Opus
59.5
Llama 3 70B
56.5
Claude 3 Opus
53
Llama 3 70B
50
Claude 3 Opus
60
Llama 3 70B
57
Claude 3 Opus
58
Llama 3 70B
55
Claude 3 Opus scores higher overall with 51 vs 48, a difference of 3 points across all benchmarks.
Claude 3 Opus leads in knowledge tasks with an average score of 59.5 vs 56.5.
Claude 3 Opus leads in coding with an average score of 53 vs 50.
Claude 3 Opus leads in math with an average score of 60 vs 57.
Claude 3 Opus leads in reasoning with an average score of 58 vs 55.