What are the hardware requirements for Pixel?

Pixel requires approximately 4.4 GB of VRAM and supports a context window of 8K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Pixel and how large is it?

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

Where can I download or evaluate Pixel?

Pixel is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Pixel by NicklasMatzulla — VRAM 4.4GB, 8K context

Name: Pixel
Author: NicklasMatzulla

Pixel is an open-source language model by NicklasMatzulla. Features: 7b LLM, VRAM: 4.4GB, Context: 8K, License: apache-2.0, Quantized, Instruction-Based, LLM Explorer Score: 0.13.

Base model:mistralai/mistral-7... Base model:quantized:mistralai... Conversational En Endpoints compatible Gguf Instruct Mistral Quantized Region:us Unsloth

Model Card on HF 🤗: https://huggingface.co/NicklasMatzulla/Pixel

Pixel Parameters and Internals

Model Type

text-generation-inference, transformers

Additional Notes

This mistral model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.

LLM Name	Pixel
Repository 🤗	https://huggingface.co/NicklasMatzulla/Pixel
Base Model(s)	mistralai/Mistral-7B-Instruct-v0.3 mistralai/Mistral-7B-Instruct-v0.3
Model Size	7b
Required VRAM	4.4 GB
Updated	2025-02-01
Maintainer	NicklasMatzulla
Model Type	mistral
Instruction-Based	Yes
Model Files	14.5 GB 4.4 GB
Supported Languages	en
GGUF Quantization	Yes
Quantization Type	gguf
Model Architecture	AutoModel
License	apache-2.0
Context Length	8192
Model Max Length	8192
Transformers Version	4.41.2
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|eot_id\|>
Vocabulary Size	128256
Torch Data Type	bfloat16

Rank the Pixel capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

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

Best Alternatives to Pixel

Best Alternatives	Context / RAM	Downloads	Likes
...ral 7B Instruct V0.2 Llamafile	0K / 14.5 GB	8850	25
Mistral 7B Instruct V0.3 GGUF	0K / 1.6 GB	91691	147
Qwen2 7B Instruct GGUF	0K / 1.9 GB	56414	10
...hemeng Qwen Math 7b 24 1 100 1	0K / 15.2 GB	25	0
Mistral 7B Instruct V0.2 GGUF	0K / 3.1 GB	70243	511
Qwen2 7B Instruct V0.6 GGUF	0K / 4.5 GB	13522	0
Qwen2 7B Instruct V0.1 GGUF	0K / 4.5 GB	9714	0
Qwen2 7B Instruct V0.7 GGUF	0K / 4.5 GB	9530	0
Qwen2 7B Instruct V0.3 GGUF	0K / 4.5 GB	8911	1
Qwen2 7B Instruct V0.2 GGUF	0K / 4.5 GB	7986	0

Note: green Score (e.g. "73.2") means that the model is better than NicklasMatzulla/Pixel.