What are the hardware requirements for Starcoder2 7B?

Starcoder2 7B requires approximately 14.4 GB of VRAM and supports a context window of 16K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Starcoder2 7B and how large is it?

Starcoder2 7B is developed by bigcode, 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 Starcoder2 7B perform on standard benchmarks?

Starcoder2 7B has the following published scores: MMLU 41.21, HumanEval (Python) 34.09. Compare against reference models on this page or on the LLM Explorer leaderboards.

Starcoder2 7B by bigcode — VRAM 14.4GB, 16K context

Name: Starcoder2 7B
Author: bigcode

Starcoder2 7B is an open-source language model by bigcode. Features: 7b LLM, VRAM: 14.4GB, Context: 16K, License: bigcode-openrail-m, LLM Explorer Score: 0.23, Arc: 38.3, HellaSwag: 51.9, MMLU: 41.2, GSM8K: 25.1, HumanEval: 34.1.

Arxiv:2004.05150 Arxiv:2205.14135 Arxiv:2207.14255 Arxiv:2305.13245 Arxiv:2402.19173 Code Dataset:bigcode/the-stack-v2-t... Deploy:azure Endpoints compatible Model-index Region:us Safetensors Sharded Starcoder2 Tensorflow

Model Card on HF 🤗: https://huggingface.co/bigcode/starcoder2-7b

Starcoder2 7B Benchmarks

ARC: 38.31 vs 96.7 (so35)^-60.4%

HellaSwag: 51.91 vs 95.3 (gpt4)^-45.5%

MMLU: 41.21 vs 88.3 (so35)^-53.3%

GSM8K: 25.09 vs 96.4 (so35)^-74%

HumanEvalA: 26.1

Python Eva: 34.09 vs 92 (so35)^-62.9%

Java Eval: 29.42

Javascript: 35.35

CPP Eval: 33.63

PHP Eval: 30.58

Julia Eval: 20.42

D Eval: 15.12

Lua Eval: 30.67

R Eval: 16.72

Racket Eva: 11.58

Rust Eval: 29.62

Swift Eval: 26.06

LLME Score: 0.22955

^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 🚀

Starcoder2 7B Parameters and Internals

Model Type

text generation

Use Cases

Areas:

research, commercial applications

Applications:

software development, educational tools

Primary Use Cases:

code generation, automated coding assistance

Limitations:

Not guaranteed to generate functioning code, Might contain inefficiencies or bugs

Additional Notes

Model is not an instruction model and may not perform well with explicit commands like writing a square root computation.

Supported Languages

EN (High), multi-language (Limited)

Training Details

Data Sources:

GitHub code, Arxiv, Wikipedia

Data Volume:

3.5+ trillion tokens

Methodology:

Grouped Query Attention, a context window of 16,384 tokens, sliding window attention of 4,096 tokens, Fill-in-the-Middle objective

Hardware Used:

432 H100 GPUs

Model Architecture:

Transformer decoder with grouped-query and sliding window attention and Fill-in-the-Middle objective

Input Output

Accepted Modalities:

text

LLM Name	Starcoder2 7B
Repository 🤗	https://huggingface.co/bigcode/starcoder2-7b
Model Size	7b
Required VRAM	14.4 GB
Updated	2026-05-07
Maintainer	bigcode
Model Type	starcoder2
Model Files	4.9 GB: 1-of-3 5.0 GB: 2-of-3 4.5 GB: 3-of-3
Model Architecture	Starcoder2ForCausalLM
License	bigcode-openrail-m
Context Length	16384
Model Max Length	16384
Transformers Version	4.37.0.dev0
Tokenizer Class	GPT2Tokenizer
Vocabulary Size	49152
Torch Data Type	bfloat16
Activation Function	gelu

Quantized Models of the Starcoder2 7B

Model	Likes	Downloads	VRAM
Starcoder2 7B AWQ	1	101	4 GB
StarCoder2 7B GGUF	20	2969	2 GB

Best Alternatives to Starcoder2 7B

Best Alternatives	Context / RAM	Downloads	Likes
Dolphincoder Starcoder2 7B	16K / 14.9 GB	7	11
Starcoder2 7B Int4 Ov	16K / 3.8 GB	15	0
Jmg Starcoder2 7B 100K	16K / 14.4 GB	5	0
Dolphincoder Starcoder2 7B	16K / 14.9 GB	278	11
Starcoder2 7B Instruct	16K / 14.4 GB	70	4
Speechless Starcoder2 7B	16K / 14.4 GB	9	4
OpenCodeInterpreter SC2 7B	16K / 14.9 GB	29	14
Starcoder2 Chat	16K / 28.8 GB	13	2
...Starcoder2 7B Bnb 4bit Smashed	16K / 4.4 GB	6	0
Starcoder2 7B 4bit	16K / 4.4 GB	274	2

Rank the Starcoder2 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

Starcoder2 7B by bigcode