What are the hardware requirements for CodeShell 7B?

CodeShell 7B requires approximately 15.4 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed CodeShell 7B and how large is it?

CodeShell 7B is developed by WisdomShell, a model with 7b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does CodeShell 7B perform on standard benchmarks?

CodeShell 7B has the following published scores: HumanEval (Python) 34.32. Compare against reference models on this page or on the LLM Explorer leaderboards.

CodeShell 7B by WisdomShell — VRAM 15.4GB

Name: CodeShell 7B
Author: WisdomShell

CodeShell 7B is an open-source language model by WisdomShell. Features: 7b LLM, VRAM: 15.4GB, LLM Explorer Score: 0.11, HumanEval: 34.3.

Codeshell Custom code En Kclgpt Openbankai Pku-kcl Pytorch Region:us Safetensors Sharded Tensorflow Wisdomshell Zh

Model Card on HF 🤗: https://huggingface.co/WisdomShell/CodeShell-7B

CodeShell 7B Benchmarks

HumanEvalA: 24.74

Python Eva: 34.32 vs 92 (so35)^-62.7%

Java Eval: 30.43

Javascript: 33.17

CPP Eval: 28.21

PHP Eval: 30.87

Julia Eval: 22.08

LLME Score: 0.10587

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

🌟 Advertise your project 🚀

CodeShell 7B Parameters and Internals

Model Type

multi-language code LLM

Supported Languages

zh (comprehensive), en (comprehensive)

Training Details

Data Volume:

500 billion tokens

Context Length:

8194

Model Architecture:

GPT-2 with Grouped-Query Attention and RoPE relative position encoding

LLM Name	CodeShell 7B
Repository 🤗	https://huggingface.co/WisdomShell/CodeShell-7B
Model Size	7b
Required VRAM	15.4 GB
Updated	2026-04-10
Maintainer	WisdomShell
Model Type	kclgpt
Model Files	10.0 GB: 1-of-2 5.4 GB: 2-of-2
Supported Languages	zh en
Model Architecture	CodeShellForCausalLM
Model Max Length	8192
Transformers Version	4.31.0
Tokenizer Class	GPT2Tokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	70144
Torch Data Type	bfloat16
Activation Function	gelu_pytorch_tanh

Best Alternatives to CodeShell 7B

Best Alternatives	Context / RAM	Downloads	Likes
Shell 7B	0K / 15.4 GB	21	0
Shell 7B Chat	0K / 15.3 GB	5	1
CodeShell 7B Chat	0K / 15.4 GB	31	25

Note: green Score (e.g. "73.2") means that the model is better than WisdomShell/CodeShell-7B.

Rank the CodeShell 7B Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

What open-source LLMs or SLMs are you in search of? 53972 in total.

Email us: info@extractum.io. Our Privacy Policy | Terms and Conditions | Suggest an improvement.

Our Social Media →

Original data from HuggingFace, OpenCompass and various public git repos.

Check out Ag3ntum — our secure, self-hosted AI agent for server management.

Release v20260328a

Support LLM Explorer

CodeShell 7B by WisdomShell

» All LLMs » WisdomShell » CodeShell 7B URL Share it on

CodeShell 7B Benchmarks

CodeShell 7B Parameters and Internals

Best Alternatives to CodeShell 7B

Rank the CodeShell 7B Capabilities

What open-source LLMs or SLMs are you in search of? 53972 in total.