Compare AI models

Specifications
Parameters
16B9B
Context window
32k128k
Benchmarks
HumanEval
57.3%82.3%
Overview
CompanyDeepSeekZ.ai
Release dateMay 16 2024Jul 5 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

CodeGeeX4-ALL-9B leads DeepSeek V2 Lite on 1 of the 1 benchmark they both report (HumanEval). DeepSeek V2 Lite shipped 50 days before CodeGeeX4-ALL-9B, so benchmark comparisons should account for the intervening progress.

DeepSeek V2 Lite has 16B parameters, while CodeGeeX4-ALL-9B has 9B. Context windows are 32k (DeepSeek V2 Lite) vs 128k (CodeGeeX4-ALL-9B).

On HumanEval, CodeGeeX4-ALL-9B leads at 82.3% vs DeepSeek V2 Lite at 57.3%.