Alibaba/Dense

Qwen3.5-35B-A3B

Name: Qwen3.5-35B-A3B
Author: Alibaba

> This repository contains model weights and configuration files for the post-trained model in the Hugging Face Transformers format. > These artifacts are compatible with Hugging Face Transformers, vLLM, SGLang, KTransformers, etc.

chatThinkingTool Use

36B

Parameters

256K

Context length

Benchmarks

Quantizations

2.5M

HF downloads

Architecture

Dense

Released

2026-02-24

Layers

KV Heads

Head Dim

128

Family

qwen

Quantization Options

Quant	Bits	VRAM	Quality
IQ3_XXS	3.25	15.1 GB	low
IQ3_XS	3.5	16.2 GB	low
Q3_K_S	3.64	16.9 GB	low
IQ3_M	3.76	17.4 GB	low
Q3_K_M	4	18.5 GB	low
Q3_K_L	4.3	19.8 GB	moderate
IQ4_XS	4.46	20.6 GB	moderate
Q4_K_S	4.67	21.5 GB	moderate
Q4_K_M	4.89	22.5 GB	good
Q5_K_S	5.57	25.6 GB	good
Q5_K_M	5.7	26.1 GB	good
Q6_K	6.56	30.0 GB	excellent
Q8_0	8.5	38.7 GB	lossless
FP16	16	72.5 GB	lossless

Select your GPU above to see speed estimates and compatibility for each quantization.

▸ READY TO RUN THIS?RENT BY THE HOUR

RENT A GPU AND RUN QWEN3.5-35B-A3B NOW

Rent on RunPod →Or Vast.ai →

Spin up an A100 / H100 / 4090 in ~60s. Pay by the second. Cancel anytime.

Community Ratings

Loading ratings...

Benchmarks (16)

τ²-Bench89.2

GPQA Diamond84.5

IFEval78.9

IFBench72.5

AA Long Context62.7

MATH59.7

BBH58.3

MMLU-PRO52.9

SciCode37.7

AA Intelligence37.1

BigCodeBench32.3

AA Coding30.3

Terminal-Bench26.5

HLE19.7

MUSR19.1

GPQA15.2

Run this model

▸Easiest way to get started·Beginners

DOCS ↗

curl -fsSL https://ollama.com/install.sh | sh

$ollama run qwen3:36.0b-instruct-q4_K_M

Downloads and runs automatically. Add --verbose for speed stats.

▸ SETUP GUIDE

Auto-setup with fitmyllm CLI

Detects your GPU, recommends the best model, downloads it, and starts chatting — zero config. Benchmarks your speed and contributes anonymous data to improve predictions.

pip install fitmyllmthen run fitmyllmLearn more

Auto-detect GPULive tok/s in chatSpeed benchmarks9 inference engines

HuggingFace Ollama Library GGUF Downloads Build Hardware

GPUs that can run this model

At Q4_K_M quantization. Sorted by minimum VRAM.

NVIDIA RTX 4090

24 GB VRAM • 1008 GB/s

NVIDIA

$1599