How much VRAM does Qwen3.5 122B A10B need?

Qwen3.5 122B A10B (122B parameters) requires approximately 88.0 GB of memory with Q3_K_M quantization.

What is the best quantization for Qwen3.5 122B A10B?

The recommended quantization for Qwen3.5 122B A10B is Q3_K_M, which balances quality and memory efficiency.

Can it run?

Yes, NVIDIA H100 80GB can run Qwen3.5 122B A10B with a C grade (Very compromised). Expected decode speed: 34.6 tok/s.

CUsable

Very compromised

Using Q3_K_M in Ollama

Capabilities:

Fit status

Very compromised

Decode

34.6 tok/s

TTFT

5594 ms

Safe context

15K

Memory

88.0 GB / 80.0 GB

Offload

10%

Weights59.8 GB

KV Cache19.1 GB

Runtime1.2 GB

Headroom8.0 GB

Workload	Grade	Fit	Decode	TTFT	Context
Agentic Coding	F	Too heavy	43.8 tok/s	6433 ms	24K
Chat	C	Runs with offload	43.8 tok/s	2412 ms	8K
Coding	C	Very compromised	34.6 tok/s	5594 ms	15K
RAG	F	Too heavy	43.8 tok/s	8041 ms	24K
Reasoning	C	Very compromised (needs ~5.5 GB host RAM)	41.8 tok/s	5477 ms	15K

How Qwen3.5 122B A10B (122B params) fits at each quantization level on NVIDIA H100 80GB (80.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	47.6 GB	Low	C42
Q3_K_SBest for your GPU	3	59.8 GB	Low	C45

HuggingFace

huggingface-cli download hf-unsloth--qwen3-5-122b-a10b-gguf

Upgrade options

MacBook Pro M3 Max 128GBBest value

C3.7 tok/s decode

~$2,499 MSRP

AMD Instinct MI350X 288GBBudget pick

C90.9 tok/s decode

~$8,000 MSRP

NVIDIA H200 141GBNVIDIA upgrade