Compare AI models

Specifications
Parameters
141B110B
Context window
64k32k
API pricing
Cheapest input
$2.00Mistral—
Cheapest output
$6.00Mistral—
$700Mistral · zdr—
Benchmarks
HumanEval
45.1%52.4%
HellaSwag
88.5%87.5%
MMLU
77.75%80.4%
GSM8K
78.6%85.4%
Overview
CompanyMistralQwen
Release dateApr 10 2024Apr 25 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Qwen1.5-110B leads Mixtral 8×22B on 3 of the 4 benchmarks they both report (HumanEval, HellaSwag, MMLU, GSM8K). Mixtral 8×22B shipped 15 days before Qwen1.5-110B, so benchmark comparisons should account for the intervening progress.

Mixtral 8×22B has 141B parameters, while Qwen1.5-110B has 110B. Context windows are 64k (Mixtral 8×22B) vs 32k (Qwen1.5-110B).

On HumanEval, Qwen1.5-110B leads at 52.4% vs Mixtral 8×22B at 45.1%. On HellaSwag, Mixtral 8×22B leads at 88.5% vs Qwen1.5-110B at 87.5%. On MMLU, Qwen1.5-110B leads at 80.4% vs Mixtral 8×22B at 77.75%. On GSM8K, Qwen1.5-110B leads at 85.4% vs Mixtral 8×22B at 78.6%.