Compare AI models

Specifications
Parameters
—14.3B
Context window
200k32k
Benchmarks
HumanEval
75.9%34.2%
MMLU
76.7%62.5%
GSM8K
88.9%61.5%
Overview
CompanyAnthropicQwen
Release dateMar 4 2024Mar 28 2024
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Claude 3 Haiku leads Qwen1.5-MoE on 3 of the 3 benchmarks they both report (HumanEval, MMLU, GSM8K). Claude 3 Haiku shipped 24 days before Qwen1.5-MoE, so benchmark comparisons should account for the intervening progress.

Context windows are 200k (Claude 3 Haiku) vs 32k (Qwen1.5-MoE). Claude 3 Haiku is closed, while Qwen1.5-MoE is open weight.

On HumanEval, Claude 3 Haiku leads at 75.9% vs Qwen1.5-MoE at 34.2%. On MMLU, Claude 3 Haiku leads at 76.7% vs Qwen1.5-MoE at 62.5%. On GSM8K, Claude 3 Haiku leads at 88.9% vs Qwen1.5-MoE at 61.5%.