← All comparisons

Llama 3.3 Instruct 70B vs Claude 4.5 Haiku (Reasoning)

Meta vs Anthropic — side-by-side benchmark comparison

Llama 3.3 Instruct 70BClaude 4.5 Haiku (Reasoning)
Intelligence Index14.537.1
Coding Index10.732.6
Math Index7.783.7
Output speed (tok/s)88.1142.2
Blended price ($/1M)$0.62$2.19
Time to first token (s)0.59s10.48s
aime30.0%
aime 257.7%83.7%
artificial analysis coding index10.7032.60
artificial analysis intelligence index14.5037.10
artificial analysis math index7.7083.70
gpqa49.8%67.2%
hle4.0%9.7%
ifbench47.1%54.3%
lcr15.0%70.3%
livecodebench28.8%61.5%
math 50077.3%
mmlu pro71.3%76.0%
scicode26.0%43.3%
tau226.6%54.7%
terminalbench hard3.0%27.3%

Benchmark data from Artificial Analysis.