Compare AI models

Specifications
Parameters
—671B
Context window
200k128k
Benchmarks
SWE-Bench Verified
7.2%49.2%
GPQA Diamond
33.3%71.5%
MMLU
76.7%90.8%
Overview
CompanyAnthropicDeepSeek
Release dateMar 4 2024Jan 20 2025
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek R1 leads Claude 3 Haiku on 3 of the 3 benchmarks they both report (SWE-Bench Verified, GPQA Diamond, MMLU). Claude 3 Haiku shipped 322 days before DeepSeek R1, so benchmark comparisons should account for the intervening progress.

Context windows are 200k (Claude 3 Haiku) vs 128k (DeepSeek R1). Claude 3 Haiku is closed, while DeepSeek R1 is open weight.

On SWE-Bench Verified, DeepSeek R1 leads at 49.2% vs Claude 3 Haiku at 7.2%. On GPQA Diamond, DeepSeek R1 leads at 71.5% vs Claude 3 Haiku at 33.3%. On MMLU, DeepSeek R1 leads at 90.8% vs Claude 3 Haiku at 76.7%.