ModelsBuildsConfigureGuidesMachinesMy Rig
Build My Rig
Build My Rig
Loading

Build it locally. We'll sort the hardware.

ModelsGPUsBuildsMachinesWorkflowsRecommendConfigureCompareGuidesAboutOpen Data
Dark mode active

New models and GPUs, straight to your inbox

Hardware updates only. Unsubscribe anytime. Privacy

Ask AI for a summary about OwnRig

Trademark Notice: NVIDIA, GeForce, and RTX are trademarks of NVIDIA Corporation. AMD and Radeon are trademarks of Advanced Micro Devices, Inc. Apple, Mac, and Apple Silicon are trademarks of Apple Inc. All other product names, logos, and brands are property of their respective owners. AI model names (Llama, Gemma, Mistral, Qwen, etc.) are trademarks of their respective creators. Use of these names and logos is for identification purposes only and does not imply endorsement.

Independence & Affiliates: OwnRig is an independent resource. We are not affiliated with, endorsed by, or sponsored by any hardware manufacturer, AI model provider, or retailer. Our recommendations are based on technical merit and community benchmarks. Some links on this site are affiliate links. If you purchase through them, we may earn a small commission at no extra cost to you. This does not influence our recommendations.

Data Accuracy: Performance figures are estimates based on community benchmarks and may vary by configuration, driver version, and software. Prices are approximate US retail as of March 2026 and may vary by retailer and region. VRAM requirements are calculated from model parameters with overhead estimates. Always verify specifications with manufacturer documentation before purchasing.

© 2026 OwnRig. All rights reserved.

Privacy
Qwen
  1. Home
  2. /Models
  3. /Qwen3.6-35B-A3B
Qwen
ChatCodingAI codingReasoningMulti-purpose35B
Chat

Qwen3.6-35B-A3B

Qwen · Apache 2.0

Mixture of Experts: 35B total parameters, 3B active per token.

MoE with 35B total parameters and ~3B active per token. Local stacks still load the full expert pool, so VRAM tracks total size. Q4_K_M needs ~22GB at practical context (RTX 4090 / 24GB class sweet spot). Q3_K_M (~16GB) fits 16GB cards at default context with minimal KV headroom; long 262K runs need KV cache quantization. 8GB cards cannot run this model without heavy CPU offload. Apache 2.0.

Parameters
35B
Architecture
MoE (3B active)
Context
262,144 tokens
Released
2026-04-01
Engines
llama.cpp, ollama, vLLM
Builder Tools
Continue, LM Studio, Open WebUI

Parameters

35B

VRAM

25 GB

Context

256K

Formats

4

GPUs

12

Qwen3.6-35B-A3B (35B) requires 25 GB VRAM at recommended quality (Q5_K_M). At efficient quality (Q4_K_M), it fits in 22 GB VRAM, making it compatible with the AMD Radeon RX 9060 XT 16GB. On NVIDIA RTX PRO 6000 Blackwell, expect approximately 278 tok/s at Q5_K_M. For the best experience, High-End Home AI Server ($3,842) is recommended.

Source: OwnRig methodology

VRAM (Recommended)

25 GB

Quantization

Q5_K_M

File Size

24.9 GB

Max Context

256K tokens

Primary Use

Chat

Memory

VRAM Requirements

QualityQuantizationVRAMFile Size
fullQ8_037 GB36.9 GB
recommendedQ5_K_M25 GB24.9 GB
efficientQ4_K_M22 GB22.1 GB
compressedQ3_K_M16 GB16.6 GB
Scaling

Context Length Impact

KV cache VRAM at Q5_K_M quality. Longer context = more memory.

ContextKV CacheTotal VRAM
2K205 MB25.2 GBexceeds 24 GB
4K512 MB25.5 GBexceeds 24 GB
8K922 MB25.9 GBexceeds 24 GB
16K1.8 GB26.8 GBexceeds 24 GB
32K3.7 GB28.7 GBexceeds 24 GB
64K7.4 GB32.4 GBexceeds 24 GB
128K14.7 GB39.7 GBexceeds 24 GB

Compatible GPUs

12 devices
NVIDIA Grace Blackwell Ultra GB300Q5_K_M145 tok/sExcellent
NVIDIA RTX PRO 6000 BlackwellQ5_K_M278 tok/sExcellent
NVIDIA RTX PRO 6000 Blackwell Max-QQ5_K_M256 tok/sExcellent
Apple M4 Ultra (192GB)Q5_K_M25 tok/sGood
NVIDIA GeForce RTX 4090Q4_K_M25 tok/sGood
AMD Radeon RX 7900 XTXQ4_K_M22 tok/sGood
Apple M4 Max (128GB Unified)Q5_K_M17 tok/sAcceptable
Apple M4 Max (64GB Unified)Q5_K_M14 tok/sAcceptable
AMD Radeon Pro W7900Q5_K_M15 tok/sAcceptable
Apple M3 Pro (18GB Unified)Q3_K_M3 tok/sMarginal
AMD Radeon RX 9070Q3_K_M–Marginal
AMD Radeon RX 9060 XT 16GBQ3_K_M–Marginal

Showing 12 of 12 entries

Build

Builder Context

Qwen3.6-35B-A3B is commonly used with Continue, LM Studio, Open WebUI. For an AI coding workflow, pair it with an embedding model like nomic-embed-text for local RAG.

FAQ

Frequently Asked Questions

How much VRAM does Qwen3.6-35B-A3B need?
Qwen3.6-35B-A3B requires 25 GB VRAM at recommended quality (Q5_K_M). At lower quality settings, it can fit in as little as 16 GB.
What is the best GPU for Qwen3.6-35B-A3B?
The NVIDIA RTX PRO 6000 Blackwell delivers the best performance for Qwen3.6-35B-A3B, achieving 278 tok/s at Q5_K_M with an excellent rating.
What quantization should I use for Qwen3.6-35B-A3B?
For the best quality, use Q5_K_M (25 GB VRAM). If your GPU has limited VRAM, Q3_K_M (16 GB) is the most efficient option with acceptable quality.
Is Qwen3.6-35B-A3B good for coding?
Yes. Qwen3.6-35B-A3B is used with Continue, LM Studio, Open WebUI for local AI coding. For the best coding experience, pair it with an embedding model for local RAG.
All models

Data confidence: community. Source

VRAM requirements are calculated from model parameters and may vary by inference engine, context length, and batch size. Performance estimates are based on community benchmarks and should be verified for your specific configuration.Qwen is a trademark of its respective owner. OwnRig is not affiliated with or endorsed by the model creator.