Alibaba Cloud122B (10B active)~74 GB VRAM at Q4_K_M

Qwen 3.5 122B-A10B — VRAM, speed & local setup

Written by Jakub Rusinowski · Last updated

Large MoE model with 122B total parameters (256 routed + 1 shared expert) and 10B active per token. Q4_K_M is ~74 GB — needs 2x24GB+ GPUs, a single 80GB GPU, or a Mac Studio with 96GB+ unified memory. Delivers near-frontier performance for long-context enterprise tasks.

Qwen 3.5 122B-A10B needs about 74 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache. On Apple Silicon that figure comes out of unified memory.

Logic93
Creative91
Coding93

VRAM and speed by quantization

Quoted against NVIDIA RTX 4090 (24 GB). Weights plus overhead only: this model's architecture is unpublished, so no KV cache is added.

QuantVRAMSpeed (est.)Fit
Q2_K
2.63 bpw
40.9 GB~18 tok/sOffload
Q3_K_M
3.41 bpw
52.8 GB~15 tok/sOffload
Q4_K_M
4.83 bpw
74.5 GB—Too big
Q5_K_M
5.67 bpw
87.3 GB—Too big
Q6_K
6.56 bpw
100.8 GB—Too big
Q8_0
8.50 bpw
130.4 GB—Too big
F16
16.00 bpw
244.8 GB—Too big

Black marker = usable memory on the NVIDIA RTX 4090 (24 GB). Estimates from the memory-bandwidth roofline on the methodology page. Qwen 3.5 122B-A10B VRAM calculator →

Get Qwen 3.5 122B-A10B running

The cheapest catalogued GPU that runs Qwen 3.5 122B-A10B is the NVIDIA DGX Spark (128 GB).

Buy this hardware NVIDIA DGX Spark (128GB) — 128 GB VRAM · 150 W board powerDeploy in the cloud now NVIDIA A100 80GB on RunPod — from $1.39/hr · rate checked 2026-07

or compare on Vast.ai from $0.77/hr (typical low · varies)

As an Amazon Associate we earn from qualifying purchases. Cloud GPU links are referral links — we may earn a commission at no extra cost to you.

Affiliate disclosure: Some links on this page are affiliate links — if you buy through them, LLM Configurator may earn a commission at no extra cost to you. As an Amazon Associate, LLM Configurator earns from qualifying purchases.
NVIDIA DGX Spark (128GB)
128 GB VRAM · 150 W board power
2026 prices are volatile — check the current listing.

How to run Qwen 3.5 122B-A10B

Install Ollama, then run:

ollama run qwen3.5:122b
Weights on Hugging Face: Qwen/Qwen3.5-122B-A10B-Instruct ↗

Specifications

Curated — A hand-written entry from before this catalogue recorded its sources. The figures are long-standing but their provenance is not on file.

Parameters
122 Billion (10B active)
Context window
262,144
Architecture
Hybrid Gated DeltaNet + MoE
Provider
Alibaba Cloud
Licence
Apache 2.0
Practical deployment
Workstation (large unified memory or pro GPU)
Specified at
Q4_K_M
System RAM
128 GB
Record updated
2026-02-24
LicenceApache-2.0Commercial use permitted

Commercial use permitted. No usage restrictions beyond attribution.

Quality and use cases

Scores as published by the model’s authors or an independent evaluator — quality, not throughput, and not measured by us.

Best forenterpriselong contextresearchmulti gpu

Can I run Qwen 3.5 122B-A10B on my GPU?

Other Qwen 3.5 sizes

Qwen 3.5 122B-A10B — frequently asked questions

How much VRAM does Qwen 3.5 122B-A10B need?

About 74 GB at Q4_K_M — quantized weights plus framework overhead, before any KV cache. The cache grows with context length and is added on top; the table above folds it in. Apple Silicon counts unified memory toward the same figure.

Does Qwen 3.5 122B-A10B run on an RTX 4090 (24 GB)?

No. Qwen 3.5 122B-A10B needs about 74 GB at Q4_K_M, more than a single RTX 4090's 24 GB. It needs a larger card, several GPUs, or Apple Silicon with enough unified memory — or it runs with part of the weights offloaded to system RAM, which is much slower.

How do I run Qwen 3.5 122B-A10B locally?

Install Ollama and run `ollama run qwen3.5:122b`. That pulls the weights and starts a local OpenAI-compatible endpoint; after the download nothing leaves the machine.

What other sizes does Qwen 3.5 come in?

Qwen 3.5 0.8B (1 GB), Qwen 3.5 2B (2 GB), Qwen 3.5 4B (3 GB), Qwen 3.5 9B (6 GB), Qwen 3.5 27B (17 GB), Qwen 3.5 35B-A3B (22 GB), Qwen 3.5 122B-A10B (74 GB), Qwen 3.5 397B-A17B (240 GB). Every size shares the family's training and licence; the larger ones score higher and need proportionally more memory.