What are the hardware requirements for Cpt Round4?

Cpt Round4 requires approximately 16 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Cpt Round4 and how large is it?

Cpt Round4 is developed by Puujeeeeeeeeeeee, a model with 8b 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 Cpt Round4?

Cpt Round4 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Cpt Round4 by Puujeeeeeeeeeeee — VRAM 16GB

Name: Cpt Round4
Author: Puujeeeeeeeeeeee

Cpt Round4 is an open-source language model by Puujeeeeeeeeeeee. Features: 8b LLM, VRAM: 16GB, License: apache-2.0, LLM Explorer Score: 0.25.

Base model:finetune:puujeeeeee... Base model:puujeeeeeeeeeeee/cp... En Endpoints compatible Gemma4 Image-text-to-text Region:us Safetensors Unsloth

Model Card on HF 🤗: https://huggingface.co/Puujeeeeeeeeeeee/cpt-round4

Cpt Round4 Parameters and Internals

LLM Name	Cpt Round4
Repository 🤗	https://huggingface.co/Puujeeeeeeeeeeee/cpt-round4
Model Name	Puujeeeeeeeeeeee/cpt-round3
Base Model(s)	Puujeeeeeeeeeeee/cpt-round3 Puujeeeeeeeeeeee/cpt-round3
Model Size	8b
Required VRAM	16 GB
Updated	2026-08-09
Maintainer	Puujeeeeeeeeeeee
Model Type	gemma4
Model Files	16.0 GB
Supported Languages	en
Model Architecture	Gemma4ForConditionalGeneration
License	apache-2.0
Model Max Length	131072
Tokenizer Class	GemmaTokenizer
Padding Token	<pad>
Torch Data Type	bfloat16

Rank the Cpt Round4 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 Cpt Round4

Best Alternatives	Context / RAM	Downloads	Likes
Gemma 4 E4B It	0K / 16 GB	4027807	1785
Gemma 4 E4B It	0K / 16 GB	229908	23
Gemma 4 E4B	0K / 16 GB	268109	386
...DIA Nemotron 3.5 Bolt E4B BF16	0K / 15.9 GB	3418	56
...4B It Weight Space Programming	0K / 16 GB	65	3
...otron 3.5 Bolt E4B Artemis FP8	0K / 2 GB	61	3
...Nemotron 3.5 Bolt E4B MLX OQ8e	0K / 8.9 GB	150	1
Vishwamai Stage1	0K / 16 GB	472	0
Lite OmniJev	0K / 32.1 GB	184	1
Gemma4 E4B FACEBOOK E2 BF16	0K / 16 GB	30	0

Note: green Score (e.g. "73.2") means that the model is better than Puujeeeeeeeeeeee/cpt-round4.