Compare AI models

Specifications
Parameters
7B/13B/70B110B
Context window
4k32k
Benchmarks
HumanEval
29.9%52.4%
MATH
13.5%49.6%
HellaSwag
85.3%87.5%
MMLU
68.9%80.4%
GSM8K
56.8%85.4%
Overview
CompanyMetaQwen
Release dateJul 18 2023Apr 25 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Qwen1.5-110B leads LLaMA 2 on 5 of the 5 benchmarks they both report. LLaMA 2 shipped 282 days before Qwen1.5-110B, so benchmark comparisons should account for the intervening progress.

LLaMA 2 has 7B/13B/70B parameters, while Qwen1.5-110B has 110B. Context windows are 4k (LLaMA 2) vs 32k (Qwen1.5-110B).

On HumanEval, Qwen1.5-110B leads at 52.4% vs LLaMA 2 at 29.9%. On MATH, Qwen1.5-110B leads at 49.6% vs LLaMA 2 at 13.5%. On HellaSwag, Qwen1.5-110B leads at 87.5% vs LLaMA 2 at 85.3%. On MMLU, Qwen1.5-110B leads at 80.4% vs LLaMA 2 at 68.9%. On GSM8K, Qwen1.5-110B leads at 85.4% vs LLaMA 2 at 56.8%.