Compare AI models

Specifications
Parameters
70B6B
Context window
16k32k
Benchmarks
HumanEval
53%9.8%
Overview
CompanyMetaZ.ai
Release dateJan 29 2024Jun 25 2023
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Code Llama 70B leads ChatGLM2-6B on 1 of the 1 benchmark they both report (HumanEval). ChatGLM2-6B shipped 218 days before Code Llama 70B, so benchmark comparisons should account for the intervening progress.

Code Llama 70B has 70B parameters, while ChatGLM2-6B has 6B. Context windows are 16k (Code Llama 70B) vs 32k (ChatGLM2-6B).

On HumanEval, Code Llama 70B leads at 53% vs ChatGLM2-6B at 9.8%.