← All comparisons

Olmo 3.1 32B Instruct vs DeepSeek V3.2 (Non-reasoning)

Allen Institute for AI vs DeepSeek — side-by-side benchmark comparison

Olmo 3.1 32B InstructDeepSeek V3.2 (Non-reasoning)
Intelligence Index12.232.1
Coding Index5.634.6
Math Index—59.0
Output speed (tok/s)0.00.0
Blended price ($/1M)$0.00$0.78
Time to first token (s)0.00s0.00s
aime——
aime 25—59.0%
artificial analysis coding index5.6034.60
artificial analysis intelligence index12.2032.10
artificial analysis math index—59.00
gpqa53.9%75.1%
hle4.9%10.5%
ifbench39.2%49.0%
lcr0.0%39.0%
livecodebench—59.3%
math 500——
mmlu pro—83.7%
scicode16.7%38.7%
tau221.3%78.9%
terminalbench hard0.0%32.6%

Benchmark data from Artificial Analysis.