Compare AI models

Specifications
Parameters
—236B
Context window
200k128k
Benchmarks
SWE-Bench Verified
40.6%19%
HumanEval
88.1%87.2%
Overview
CompanyAnthropicDeepSeek
Release dateOct 22 2024Jul 24 2024
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Claude 3.5 Haiku leads DeepSeek Coder V2-0724 on 2 of the 2 benchmarks they both report (SWE-Bench Verified, HumanEval). DeepSeek Coder V2-0724 shipped 90 days before Claude 3.5 Haiku, so benchmark comparisons should account for the intervening progress.

Context windows are 200k (Claude 3.5 Haiku) vs 128k (DeepSeek Coder V2-0724). Claude 3.5 Haiku is closed, while DeepSeek Coder V2-0724 is open weight.

On SWE-Bench Verified, Claude 3.5 Haiku leads at 40.6% vs DeepSeek Coder V2-0724 at 19%. On HumanEval, Claude 3.5 Haiku leads at 88.1% vs DeepSeek Coder V2-0724 at 87.2%.