SAVRN
Search Contact SAVRN

Open-weight model · Image and text to text

Qwen3.6-35B-A3B-NVFP4

by Unsloth AI unsloth/Qwen3.6-35B-A3B-NVFP4

1.56x faster throughput than other NVFP4 quants. This is an Unsloth NVFP4 quantized checkpoint calibrated on a mixture of our Unsloth dataset + UltraChat dataset. Works on a 32GB VRAM GPU. Benchmarks on 1xB200 128 concurrency.

Parameters24.6B
Context262,144
Weights26.5 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads1.4M

Runs On

What it takes to serve Qwen3.6-35B-A3B-NVFP4 (24.6B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

PrecisionWeightsMemory neededCheapest setupPer hourAlso fits
16-bit 49.3 GB 59.1 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 24.6 GB 29.6 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 12.3 GB 14.8 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.

SAVRN's Notes on Qwen3.6-35B-A3B-NVFP4

Of the 256 experts in each layer, 8 fire per token, so only a fraction of the 24.6B parameters touch any one token; memory is the constraint, since every expert has to be resident. The 4-bit row reads 12.3 GB of weights and 14.8 GB needed, and the publisher puts it on a 32 GB card. The cheapest listed setup, one MI300X with 192 GB at $1.85 an hour on-demand, leaves most of the card for a 262,144-token context and for image inputs.

The license is Apache 2.0, so commercial use and modification are permitted with notices retained. Before you commit, note that this is an NVFP4 quantization of Qwen/Qwen3.6-35B-A3B published by Unsloth AI, calibrated on the publisher's own dataset mixed with UltraChat, and that the publisher serves it through vLLM; no host prices for it appear in the SAVRN Index.

SAVRN Research, 2026-09-18

Model Card

By Unsloth AI, published under apache-2.0, revision 739af1e7aac3.

Read our How to Run Qwen3.6 NVFP4 Guide!

Read the full model card (3,434 words)

Configuration

Architecture
Qwen3_5MoeForConditionalGeneration
Context length (tokens)
262,144
Layers
40
Hidden size
2,048
Attention heads
16
Key/value heads
2
Head dimension
256
Vocabulary size
248,320
Experts
256
Experts active per token
8
Model type
qwen3_5_moe
Quantization
compressed-tensors

Identity and Version

Repository
unsloth/Qwen3.6-35B-A3B-NVFP4
Publisher
Unsloth AI
Task
Image and text to text
Modality
Image and text
Library
transformers
Parameters
24.6B parameters
Languages
Not stated by the source
Revision
739af1e7aac320af1682ed1e0cce369af4c5265d
First published
2026-04-23
Last updated
2026-07-12

Files and Weights

19 files, 26.5 GB in total. The weights are 6 files totalling 26.5 GB in safetensors.

Weights6 files · 26.5 GB
Configuration7 files · 12.8 MB
Tokenizer3 files · 26.7 MB
Documentation1 file · 67.4 KB
Other1 file · 7.8 KB
Repository1 file · 1.6 KB
Every file
FileTypeSizeSHA-256
model-00001-of-00006.safetensorsWeights5.0 GB ae6d6df8a4ba
model-00002-of-00006.safetensorsWeights5.0 GB 071ab0aa7b8e
model-00003-of-00006.safetensorsWeights5.0 GB 667efa338501
model-00004-of-00006.safetensorsWeights5.0 GB 6d6fc2b95dcf
model-00005-of-00006.safetensorsWeights4.8 GB 1e4d3ae27bab
model-00006-of-00006.safetensorsWeights1.7 GB 3458074b46f2
config.jsonConfiguration22.6 KB —
configuration.jsonConfiguration58 B —
generation_config.jsonConfiguration214 B —
model.safetensors.index.jsonConfiguration12.8 MB 74851913bc50
preprocessor_config.jsonConfiguration390 B —
processor_config.jsonConfiguration1.2 KB —
video_preprocessor_config.jsonConfiguration385 B —
README.mdDocumentation67.4 KB —
chat_template.jinjaOther7.8 KB —
.gitattributesRepository1.6 KB —
tokenizer.jsonTokenizer20.0 MB 1a6329cee073
tokenizer_config.jsonTokenizer1.2 KB —
vocab.jsonTokenizer6.7 MB —

License and Download

License
apache-2.0
Access
Open weights, no gate
Download size
26.5 GB
Download from Unsloth AI

Released by Unsloth AI through its official repository on Hugging Face. Read the license.

Built From

Memory Requirements

PrecisionWeights in memory
As published26.5 GB
16-bit49.3 GB
8-bit24.6 GB
4-bit12.3 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Compare Qwen3.6-35B-A3B-NVFP4

Questions About Qwen3.6-35B-A3B-NVFP4

How much GPU memory does Qwen3.6-35B-A3B-NVFP4 need?

About 59.1 GB at 16-bit and 14.8 GB at 4-bit: the weights (24.6B parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run Qwen3.6-35B-A3B-NVFP4 on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

Can I use Qwen3.6-35B-A3B-NVFP4 commercially?

Yes. Qwen3.6-35B-A3B-NVFP4 is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.

What is Qwen3.6-35B-A3B-NVFP4's context length?

262,144 tokens, from the maximum position embeddings in its published configuration.

Similar Models

Model · Image and text to text

gemma-4-26B-A4B-it

Google

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context window of up to 256K tokens and maintains multilingual support in over 140 languages. Featuring both Dense and Mixture-of-Experts (MoE) architectures, Gemma 4 is well-suited for tasks like text generation, coding, and reasoning. The models are available in five distinct sizes: E2B, E4B, 12B, 26B A4B, and 31B. Their diverse sizes make them deployable in environments ranging from…

Open weights apache-2.0 25.8B parameters 262,144 tokens transformers
View model

Model · Image and text to text

gemma-4-26B-A4B-it-AWQ-4bit

Cyankiwi

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on small models) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context window of up to 256K tokens and maintains multilingual support in over 140 languages. Featuring both Dense and Mixture-of-Experts (MoE) architectures, Gemma 4 is well-suited for tasks like text generation, coding, and reasoning. The models are available in four distinct sizes: E2B, E4B, 26B A4B, and 31B. Their diverse sizes make them deployable in environments ranging from high-end phones…

Open weights apache-2.0 25.8B parameters 262,144 tokens transformers
View model

Model · Image and text to text

Aura-Prototype-26B-A4B

EldritchLabs

This is a merge of pre-trained language models created using mergekit. This model was merged using the aura merge method. Aura is an experimental method with a live heatmap visualizer. This model took 10 hours to merge using graphv18.py The following models were included in the merge: - TheDrummer/Orion-26B-A4B-v1.1 - Gryphe/Pantheon-Reasoning-26B-A4B-1.1-V2 - electroglyph/gemma4-26b-fiction-bf16 The following YAML configuration was used to produce this model

Open weights apache-2.0 26B parameters 262,144 tokens transformers
View model

Model · Image and text to text

Qwen3.8-27B-MLX-4bit

LM Studio Community

LM Studio Community models highlights program. Highlighting new & noteworthy models by the community. Join the conversation on Discord. 4-bit quantized version of Qwen3.8-27B using MLX, optimized for Apple Silicon. Special thanks to the Apple Machine Learning Research team for creating MLX. LM Studio is not the creator, originator, or owner of any Model featured in the Community Model Program. Each Community Model is created and provided by third parties. LM Studio does not endorse, support, represent or guarantee the completeness, truthfulness, accuracy, or reliability of any Community Model. You understand that Community Models can produce content that might be offensive, harmful…

Open weights apache-2.0 27.4B parameters 262,144 tokens transformers
View model

Model · Image and text to text

Qwen3.8-27B-MLX-8bit

LM Studio Community

LM Studio Community models highlights program. Highlighting new & noteworthy models by the community. Join the conversation on Discord. 8-bit quantized version of Qwen3.8-27B using MLX, optimized for Apple Silicon. Special thanks to the Apple Machine Learning Research team for creating MLX. LM Studio is not the creator, originator, or owner of any Model featured in the Community Model Program. Each Community Model is created and provided by third parties. LM Studio does not endorse, support, represent or guarantee the completeness, truthfulness, accuracy, or reliability of any Community Model. You understand that Community Models can produce content that might be offensive, harmful…

Open weights apache-2.0 27.4B parameters 262,144 tokens transformers
View model