Compare AI models

Specifications
Parameters
—14.3B
Context window
128k32k
Benchmarks
HumanEval
83.7%34.2%
MMLU
84.7%62.5%
Overview
CompanyOpenAIQwen
Release dateNov 6 2023Mar 28 2024
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

GPT-4 Turbo leads Qwen1.5-MoE on 2 of the 2 benchmarks they both report (HumanEval, MMLU). GPT-4 Turbo shipped 143 days before Qwen1.5-MoE, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (GPT-4 Turbo) vs 32k (Qwen1.5-MoE). GPT-4 Turbo is closed, while Qwen1.5-MoE is open weight.

On HumanEval, GPT-4 Turbo leads at 83.7% vs Qwen1.5-MoE at 34.2%. On MMLU, GPT-4 Turbo leads at 84.7% vs Qwen1.5-MoE at 62.5%.