Compare AI models

Specifications
Parameters
671B109B
Context window
128k10M
API pricing
Cheapest input
—$0.08DeepInfra
Cheapest output
—$0.24DeepInfra
—$35.00DeepInfra · fp8
Benchmarks
LiveCodeBench
65.9%32.8%
GPQA Diamond
71.5%57.2%
MMLU-Pro
84%74.3%
Overview
CompanyDeepSeekMeta
Release dateJan 20 2025Apr 5 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek R1 leads LLaMA 4 Scout on 3 of the 3 benchmarks they both report (LiveCodeBench, GPQA Diamond, MMLU-Pro). DeepSeek R1 shipped 75 days before LLaMA 4 Scout, so benchmark comparisons should account for the intervening progress.

DeepSeek R1 has 671B parameters, while LLaMA 4 Scout has 109B. Context windows are 128k (DeepSeek R1) vs 10M (LLaMA 4 Scout).

On LiveCodeBench, DeepSeek R1 leads at 65.9% vs LLaMA 4 Scout at 32.8%. On GPQA Diamond, DeepSeek R1 leads at 71.5% vs LLaMA 4 Scout at 57.2%. On MMLU-Pro, DeepSeek R1 leads at 84% vs LLaMA 4 Scout at 74.3%.