What are the hardware requirements for Mnma Qwen3.8 27B Nvfp4?

Mnma Qwen3.8 27B Nvfp4 requires approximately 18.8 GB of VRAM and supports a context window of 256K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Mnma Qwen3.8 27B Nvfp4 and how large is it?

Mnma Qwen3.8 27B Nvfp4 is developed by minima-ai, a model with 27b 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 Mnma Qwen3.8 27B Nvfp4?

Mnma Qwen3.8 27B Nvfp4 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Mnma Qwen3.8 27B Nvfp4 by minima-ai — VRAM 18.8GB, 256K context

Name: Mnma Qwen3.8 27B Nvfp4
Author: minima-ai

Mnma Qwen3.8 27B Nvfp4 is an open-source language model by minima-ai. Features: 27b LLM, VRAM: 18.8GB, Context: 256K, License: apache-2.0, LLM Explorer Score: 0.32.

Arxiv:2609.04098 8-bit Base model:quantized:qwen/qwen... Base model:qwen/qwen3.8-27b Compressed-tensors Conversational Fp8-kv-cache Gated-deltanet Nvfp4 Quantized Qwen3.8 Qwen3 5 text Region:us Safetensors Vllm W4a4

Model Card on HF 🤗: https://huggingface.co/minima-ai/mnma_qwen3.8_27b_nvfp4

Mnma Qwen3.8 27B Nvfp4 Parameters and Internals

LLM Name	Mnma Qwen3.8 27b Nvfp4
Repository 🤗	https://huggingface.co/minima-ai/mnma_qwen3.8_27b_nvfp4
Base Model(s)	Qwen/Qwen3.8-27B Qwen/Qwen3.8-27B
Model Size	27b
Required VRAM	18.8 GB
Updated	2026-09-11
Maintainer	minima-ai
Model Type	qwen3_5_text
Model Files	18.8 GB
Model Architecture	Qwen3_5ForCausalLM
License	apache-2.0
Context Length	262144
Model Max Length	262144
Transformers Version	5.14.1
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	248320
Errors	replace

Rank the Mnma Qwen3.8 27B Nvfp4 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 Mnma Qwen3.8 27B Nvfp4

Best Alternatives	Context / RAM	Downloads	Likes
Qwen3.6 27B OBLITERATED	256K / 53.6 GB	15137	184
Qwen3.8 27B Code Tools Merged	256K / 53.8 GB	242	1
Qwen3.8 27B Obliterated E03	256K / 53.6 GB	708	1
Sharona Q27B R CodeSecurity V2	256K / 53.8 GB	574	1
Barcenas Qwen3.8 27B Fable	256K / 54.1 GB	869	5
...maAQUA Gridbook 13GB 5080 Vllm	256K / 13 GB	731	4
...wen3.8 P44w75 16.8B Unrepaired	256K / 2.4 GB	929	1
...2 Abliterated OptiQ 3.7bpw Mlx	256K / 13.4 GB	6727	14
...n3.8 27B PrismaScout AQUA 20GB	256K / 20.2 GB	266	3
Qwen3.8 27B Abliteratex	256K / 53.9 GB	321	6

Note: green Score (e.g. "73.2") means that the model is better than minima-ai/mnma_qwen3.8_27b_nvfp4.