← All comparisons

Llama 3.1 Nemotron Ultra 253B v1 (Reasoning) vs GPT-4.1 nano

NVIDIA vs OpenAI — side-by-side benchmark comparison

Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)GPT-4.1 nano
Intelligence Index15.013.0
Coding Index13.111.2
Math Index63.724.0
Output speed (tok/s)52.3178.9
Blended price ($/1M)$0.90$0.17
Time to first token (s)0.76s0.40s
aime74.7%23.7%
aime 2563.7%24.0%
artificial analysis coding index13.1011.20
artificial analysis intelligence index15.0013.00
artificial analysis math index63.7024.00
gpqa72.8%51.2%
hle8.1%3.9%
ifbench38.2%32.0%
lcr7.3%17.0%
livecodebench64.1%32.6%
math 50095.2%84.8%
mmlu pro82.5%65.7%
scicode34.7%25.9%
tau211.4%17.3%
terminalbench hard2.3%3.8%

Benchmark data from Artificial Analysis.