Compare AI models

Specifications
Parameters
47B110B
Context window
32k32k
Benchmarks
HumanEval
40.2%52.4%
HellaSwag
84.4%87.5%
MMLU
70.6%80.4%
GSM8K
58.4%85.4%
Overview
CompanyMistralQwen
Release dateDec 8 2023Apr 25 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Qwen1.5-110B leads Mixtral 8×7B on 4 of the 4 benchmarks they both report (HumanEval, HellaSwag, MMLU, GSM8K). Mixtral 8×7B shipped 139 days before Qwen1.5-110B, so benchmark comparisons should account for the intervening progress.

Mixtral 8×7B has 47B parameters, while Qwen1.5-110B has 110B. Context windows are 32k (Mixtral 8×7B) vs 32k (Qwen1.5-110B).

On HumanEval, Qwen1.5-110B leads at 52.4% vs Mixtral 8×7B at 40.2%. On HellaSwag, Qwen1.5-110B leads at 87.5% vs Mixtral 8×7B at 84.4%. On MMLU, Qwen1.5-110B leads at 80.4% vs Mixtral 8×7B at 70.6%. On GSM8K, Qwen1.5-110B leads at 85.4% vs Mixtral 8×7B at 58.4%.