Claude 3.5 HaikuvsDeepSeek-V3.2

Claude 3.5 Haiku
DeepSeek-V3.2
Benchmarks
Nonsense detection
BullshitBench v2
50%Best
13%
Coding
SWE-Bench Verified
40.6%
Science
GPQA Diamond
41.6%
Community preference (code)
Arena Elo (Code)
1361
Overview
CompanyAnthropicDeepSeek
Release dateOct 22 2024Dec 1 2025
AccessProprietaryOpen Weight

Which is better: Claude 3.5 Haiku or DeepSeek-V3.2?

Claude 3.5 Haiku leads DeepSeek-V3.2 on 1 of the 1 benchmark they both report (BullshitBench v2). Claude 3.5 Haiku shipped 405 days before DeepSeek-V3.2, so benchmark comparisons should account for the intervening progress.

Claude 3.5 Haiku is proprietary, while DeepSeek-V3.2 is open weight.

On BullshitBench v2, Claude 3.5 Haiku leads at 50% vs DeepSeek-V3.2 at 13%.

Frequently asked questions

Claude 3.5 Haiku was released by Anthropic on Oct 22 2024.

DeepSeek-V3.2 was released by DeepSeek on Dec 1 2025.

Claude 3.5 Haiku is a proprietary model released by Anthropic. DeepSeek-V3.2 is an open weight model released by DeepSeek.

Other comparisons