Compare AI models

Specifications
Parameters
8B/70B/405B6B
Context window
128k32k
API pricing
Cheapest input
$0.02DeepInfra—
Cheapest output
$0.04DeepInfra—
$6.00DeepInfra · fp8—
Benchmarks
HumanEval
89%9.8%
MATH
73.8%6.9%
MMLU
87.3%45.2%
GSM8K
96.8%25.9%
Overview
CompanyMetaZ.ai
Release dateJul 23 2024Jun 25 2023
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

LLaMA 3.1 leads ChatGLM2-6B on 4 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). ChatGLM2-6B shipped 394 days before LLaMA 3.1, so benchmark comparisons should account for the intervening progress.

LLaMA 3.1 has 8B/70B/405B parameters, while ChatGLM2-6B has 6B. Context windows are 128k (LLaMA 3.1) vs 32k (ChatGLM2-6B).

On HumanEval, LLaMA 3.1 leads at 89% vs ChatGLM2-6B at 9.8%. On MATH, LLaMA 3.1 leads at 73.8% vs ChatGLM2-6B at 6.9%. On MMLU, LLaMA 3.1 leads at 87.3% vs ChatGLM2-6B at 45.2%. On GSM8K, LLaMA 3.1 leads at 96.8% vs ChatGLM2-6B at 25.9%.