Compare AI models

Specifications
Parameters
671B49B
Context window
128k128k
API pricing
Cheapest input
$0.25DeepInfra—
Cheapest output
$0.95DeepInfra—
$97.50DeepInfra · fp4—
Benchmarks
LiveCodeBench
74.8%73.58%
Humanity's Last Exam · text only, no tools
15.9%7.64%
AIME 2024
93.1%87.5%
AIME 2025
88.4%82.71%
GPQA Diamond
80.1%71.97%
MMLU-Pro
84.8%79.53%
Overview
CompanyDeepSeekNVIDIA
Release dateAug 21 2025Jul 25 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek V3.1 leads Llama Nemotron Super 49B v1.5 on 6 of the 6 benchmarks they both report. Llama Nemotron Super 49B v1.5 shipped 27 days before DeepSeek V3.1, so benchmark comparisons should account for the intervening progress.

DeepSeek V3.1 has 671B parameters, while Llama Nemotron Super 49B v1.5 has 49B. Context windows are 128k (DeepSeek V3.1) vs 128k (Llama Nemotron Super 49B v1.5).

On LiveCodeBench, DeepSeek V3.1 leads at 74.8% vs Llama Nemotron Super 49B v1.5 at 73.58%. On Humanity's Last Exam · text only, no tools, DeepSeek V3.1 leads at 15.9% vs Llama Nemotron Super 49B v1.5 at 7.64%. On AIME 2024, DeepSeek V3.1 leads at 93.1% vs Llama Nemotron Super 49B v1.5 at 87.5%. On AIME 2025, DeepSeek V3.1 leads at 88.4% vs Llama Nemotron Super 49B v1.5 at 82.71%. On GPQA Diamond, DeepSeek V3.1 leads at 80.1% vs Llama Nemotron Super 49B v1.5 at 71.97%. On MMLU-Pro, DeepSeek V3.1 leads at 84.8% vs Llama Nemotron Super 49B v1.5 at 79.53%.