What are the hardware requirements for Qwen3.8 Flash Next OQ3.5e Fp16 Mtp?

Qwen3.8 Flash Next OQ3.5e Fp16 Mtp requires approximately 90.7 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Qwen3.8 Flash Next OQ3.5e Fp16 Mtp and how large is it?

Qwen3.8 Flash Next OQ3.5e Fp16 Mtp is developed by Robot-Haus, a model with 180b 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 Qwen3.8 Flash Next OQ3.5e Fp16 Mtp?

Qwen3.8 Flash Next OQ3.5e Fp16 Mtp is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Qwen3.8 Flash Next OQ3.5e Fp16 Mtp by Robot-Haus — VRAM 90.7GB

Name: Qwen3.8 Flash Next OQ3.5e Fp16 Mtp
Author: Robot-Haus

Qwen3.8 Flash Next OQ3.5e Fp16 Mtp is an open-source language model by Robot-Haus. Features: 180b LLM, VRAM: 90.7GB, License: other, Quantized, LLM Explorer Score: 0.32.

Arxiv:2502.02789 3-bit 3.5-bit Apple-silicon Base model:quantized:qwen/qwen... Base model:qwen/qwen3.8-flash-... Conversational Dataset:ai-safety-promise/safe... Dataset:cais/mmlu Dataset:google-research-datase... Dataset:huggingfaceh4/hellaswa... Dataset:huggingfaceh4/truthful... Dataset:livecodebench/livecode... Dataset:openai/openai humaneva... En Fp16 Image-text-to-text Mixture-of-experts Mlx Mlx-vlm Model-index Mtp Omlx Oq Oqe Quantized Qwen Qwen3.8 Qwen4-exp Qwen4 exp Region:us Safetensors Sharded Speculative-decoding Tensorflow Vision Vision-language

Model Card on HF 🤗: https://huggingface.co/Robot-Haus/Qwen3.8-Flash-Next-oQ3.5e-fp16-mtp

Qwen3.8 Flash Next OQ3.5e Fp16 Mtp Parameters and Internals

LLM Name	Qwen3.8 Flash Next OQ3.5e Fp16 Mtp
Repository 🤗	https://huggingface.co/Robot-Haus/Qwen3.8-Flash-Next-oQ3.5e-fp16-mtp
Base Model(s)	Qwen/Qwen3.8-Flash-Next Qwen/Qwen3.8-Flash-Next
Model Size	180b
Required VRAM	90.7 GB
Updated	2026-10-04
Maintainer	Robot-Haus
Model Type	qwen4_exp
Model Files	5.1 GB: 1-of-18 5.0 GB: 2-of-18 5.0 GB: 3-of-18 5.0 GB: 4-of-18 5.0 GB: 5-of-18 5.6 GB: 6-of-18 5.0 GB: 7-of-18 5.3 GB: 8-of-18 5.2 GB: 9-of-18 5.2 GB: 10-of-18 5.3 GB: 11-of-18 5.2 GB: 12-of-18 5.2 GB: 13-of-18 5.3 GB: 14-of-18 5.2 GB: 15-of-18 5.2 GB: 16-of-18 5.3 GB: 17-of-18 2.6 GB: 18-of-18
Supported Languages	en
Quantization Type	fp16
Model Architecture	Qwen4ExpForConditionalGeneration
License	other
Model Max Length	262144
Transformers Version	5.8.0.dev0
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Errors	replace

Rank the Qwen3.8 Flash Next OQ3.5e Fp16 Mtp 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 Qwen3.8 Flash Next OQ3.5e Fp16 Mtp

Best Alternatives	Context / RAM	Downloads	Likes
Qwen3.8 Flash Next 4bit Mtp	0K / 112.9 GB	540	3
...en3.8 Flash Next OQ4e Fp16 Mtp	0K / 107.3 GB	746	2
...wen3.8 Flash Next MLX 4bit MTP	0K / 112.8 GB	1643	5
...MLX Activation 3bit Native PLE	0K / 34.7 GB	284	1
Qwen3.8 Flash Next	0K / 129.5 GB	4810	3918
Darwin 180B RSI	0K / 119 GB	353	94
Tinfield 1	0K / 129.5 GB	577	41
Qwen3.8 Flash Next FP8	0K / 65.7 GB	2219	117
CYBER FROST 3.8 BF16	0K / 124.3 GB	935	40
Qwen3.8 Flash Next W4A16	0K / 0.5 GB	6043	22

Note: green Score (e.g. "73.2") means that the model is better than Robot-Haus/Qwen3.8-Flash-Next-oQ3.5e-fp16-mtp.