Compare AI models

Specifications
Parameters
—8B
Context window
200k—
Benchmarks
GPQA Diamond
41.6%61.1%
Overview
CompanyAnthropicDeepSeek
Release dateOct 22 2024May 29 2025
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek R1-0528 Qwen3 8B leads Claude 3.5 Haiku on 1 of the 1 benchmark they both report (GPQA Diamond). Claude 3.5 Haiku shipped 219 days before DeepSeek R1-0528 Qwen3 8B, so benchmark comparisons should account for the intervening progress.

Claude 3.5 Haiku is closed, while DeepSeek R1-0528 Qwen3 8B is open weight.

On GPQA Diamond, DeepSeek R1-0528 Qwen3 8B leads at 61.1% vs Claude 3.5 Haiku at 41.6%.