Skip to content

Qwen3 VL 235B A22B Instruct on M5 Max 128 GB: VRAM, speed and fit

Qwen3 VL 235B A22B Instruct at 2-bit runs on the M5 Max 128 GB with 44K of context at about 32 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on M5 Max 128 GB · 2-bit (Q2_K) · Ollama / LM Studio · Chat

GPU memory
123 GB
of 128 GB
Context
44K
auto
Speed
32 tok/s
20 when full
First reply
7.1 s
2K prompt

Where the memory goes

Model
82.3 GB
Context
8.1 GB
Overhead
681 MB
OS
32.0 GB
Free
5.0 GB
Memory
123 GB of 128 GB
Context
44K tokens · auto
Speed
32 tok/s · 20 when full
First reply
7.1 s for a 2K prompt
Running cost
$2.63/mo at home · $131/mo on a cloud 4× RTX A5000

Validation: Not validated

What Qwen3 VL 235B A22B Instruct on M5 Max 128 GB was checked against
CheckTargetResultStatus
Fits in GPU memory96.0 GB usable91.0 GB neededPass
Headroom5% free5.0 GB free (5%)Pass
Context for chat16K or more44KPass
Writing speed15 tok/s or more32 tok/sPass
First replyunder 3 s, 2K prompt7.1 sFails

Targets: a reply you can read along with, a first reply the use case tolerates, and 5% of memory left free. Calculated from published specs, not benchmarked.

The setup

Model
Qwen3 VL 235B A22B Instruct
Parameters
236B, 22B active
Precision
2-bit (Q2_K)
Context
44K, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
M5 Max 128 GB
GPU can use
96.0 GB of 128 GB

Worth trying

  • 8-bit cache: Room for 80K context

Running cost

At home
$2.63/mo
120 W at the wall
Renting instead
$131/mo
4× RTX A5000 at $1.08/h

4 h a day, power at $0.18 per kWh, cloud at RunPod on-demand prices as of Nov 22, 2026.

How to run it

llama.cpp
llama-server -hf unsloth/Qwen3-VL-235B-A22B-Instruct-GGUF
vLLM
vllm serve Qwen/Qwen3-VL-235B-A22B-Instruct

Qwen3 VL 235B A22B Instruct on similar machines

All machines: Qwen3 VL 235B A22B Instruct VRAM requirements. Everything that runs here: what can the M5 Max 128 GB run?

Other models that run on M5 Max 128 GB

Questions

Can Qwen3 VL 235B A22B Instruct run on M5 Max 128 GB?

Qwen3 VL 235B A22B Instruct at 2-bit runs on the M5 Max 128 GB with 44K of context at about 32 tokens a second.

Report VR-12E32TM, calculated Nov 22, 2026 from published specs, not benchmarked.