Compare AI models

Specifications
Parameters
2B/7B14.3B
Context window
8k32k
Benchmarks
HumanEval
32.3%34.2%
MMLU
64.3%62.5%
GSM8K
46.4%61.5%
Overview
CompanyGoogleQwen
Release dateFeb 21 2024Mar 28 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Qwen1.5-MoE leads Gemma 1 on 2 of the 3 benchmarks they both report (HumanEval, MMLU, GSM8K). Gemma 1 shipped 36 days before Qwen1.5-MoE, so benchmark comparisons should account for the intervening progress.

Gemma 1 has 2B/7B parameters, while Qwen1.5-MoE has 14.3B. Context windows are 8k (Gemma 1) vs 32k (Qwen1.5-MoE).

On HumanEval, Qwen1.5-MoE leads at 34.2% vs Gemma 1 at 32.3%. On MMLU, Gemma 1 leads at 64.3% vs Qwen1.5-MoE at 62.5%. On GSM8K, Qwen1.5-MoE leads at 61.5% vs Gemma 1 at 46.4%.