Qwen3.5 122B A10B needs ~73.1 GB VRAM. Mac Studio M3 Ultra 96GB has 69.1 GB. With Q2_K quantization, expect ~9 tok/s.
Operating mode
Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.
Current mode
Balanced
Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.
Select quantization to explore
206.6 GB over capacity — needs offload or smaller quantization
Fit status
Too heavy
Decode
2.0 tok/s
TTFT
96800 ms
Safe context
4K
Memory
275.7 GB / 69.1 GB
Offload
70%
It fits through host-memory offload, and offload is the main reason performance drops.
CPU or host-memory offload is active
About 10% of the working set spills out of accelerator memory, which usually hurts latency and sustained decode throughput.
Very little memory headroom
You can run the model, but there is not much room left for longer context, bigger batches, extra apps, or future model updates.
Shared-memory contention still exists
The OS, browser, and inference runtime all compete for the same physical memory pool, so real-world headroom is less forgiving than raw capacity suggests.
Remove offload with more accelerator memory
Prioritize a GPU or unified-memory tier that fits the whole model natively. Removing offload usually helps more than small compute gains.
Buy headroom, not only minimum fit
A slightly larger memory tier gives you safer context growth and makes the recommendation more future-proof.
Increase host RAM if you keep offloading
This setup may need roughly 2.6 GB of extra host RAM just for the offloaded portion, before OS and other tools.
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | D | Very compromised (needs ~6.9 GB host RAM) | 7.1 tok/s | 14838 ms | 4K |
| Coding | F | Too heavy | 6.4 tok/s | 30401 ms | 4K |
| Agentic Coding | F | Too heavy | 5.3 tok/s | 53116 ms | 4K |
| Reasoning | F | Too heavy | 6.4 tok/s | 35928 ms | 4K |
| RAG | F | Too heavy | 5.3 tok/s | 66395 ms | 4K |
How Qwen3.5 122B A10B (122B params) fits at each quantization level on Mac Studio M3 Ultra 96GB (69.1 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_KBest for your GPU | 2 | 47.6 GB | Low | C48 |
Q3_K_S | 3 | 59.8 GB | Low | F0 |
NVFP4 | 4 | 68.3 GB | Medium | F0 |
Q4_K_M | 4 | 74.4 GB | Medium | F0 |
Q5_K_M | 5 | 87.8 GB | High | F0 |
Q6_K | 6 | 100.0 GB | High | F0 |
Q8_0 | 8 | 130.5 GB | Very High | F0 |
F16 | 16 | 250.1 GB | Maximum | F0 |
Copy-paste commands to run Qwen3.5 122B A10B on your machine.
Run
docker run --rm -it ghcr.io/ggerganov/llama.cpp:full \
--hf-repo "unsloth/Qwen3.5-122B-A10B-GGUF" \
--hf-file "Qwen3.5-122B-A10B-GGUF-Q3_K_M.gguf" \
-c 4096 -ngl 99Opciones de mejora
Hace que el modelo quepa en el acelerador en lugar de seguir fuera de alcance.
Elimina el offload a memoria del sistema, que suele ser la mayor mejora individual en latencia y throughput.
~$2,499 MSRP
Hace que el modelo quepa en el acelerador en lugar de seguir fuera de alcance.
Elimina el offload a memoria del sistema, que suele ser la mayor mejora individual en latencia y throughput.
~$3,999 MSRP
Hace que el modelo quepa en el acelerador en lugar de seguir fuera de alcance.
Elimina el offload a memoria del sistema, que suele ser la mayor mejora individual en latencia y throughput.
~$3,999 MSRP
Hace que el modelo quepa en el acelerador en lugar de seguir fuera de alcance.
Elimina el offload a memoria del sistema, que suele ser la mayor mejora individual en latencia y throughput.
~$12,000 MSRP
Yes, Mac Studio M3 Ultra 96GB can run Qwen3.5 122B A10B at Q2_K quantization (Runs with offload (needs ~2.6 GB host RAM)). The recommended Q3_K_M requires 85.3 GB which exceeds available memory, but at Q2_K it needs only 73.1 GB. Expected decode speed: 9.0 tok/s.
Qwen3.5 122B A10B (122B parameters) requires approximately 85.3 GB at Q3_K_M quantization. On Mac Studio M3 Ultra 96GB, it fits at Q2_K using 73.1 GB.
The recommended quantization is Q3_K_M, but on Mac Studio M3 Ultra 96GB the best fitting quantization is Q2_K, which uses 73.1 GB.
On Mac Studio M3 Ultra 96GB, Qwen3.5 122B A10B achieves approximately 9.0 tokens per second decode speed with a time-to-first-token of 21580ms using Q2_K quantization.
For coding workloads, Qwen3.5 122B A10B on Mac Studio M3 Ultra 96GB receives a F grade with 6.4 tok/s and 4K context.
On Mac Studio M3 Ultra 96GB, Qwen3.5 122B A10B can safely use up to 11K tokens of context at Q2_K quantization. The model's official context limit is —, but available memory constrains the safe maximum.
Remove offload with more accelerator memory. Prioritize a GPU or unified-memory tier that fits the whole model natively. Removing offload usually helps more than small compute gains.
Not always. Mac Studio M3 Ultra 96GB can often fit larger models thanks to unified memory, but a discrete GPU with dedicated high-bandwidth VRAM may still decode faster once the model fits. For this combination, the important distinction is capacity versus sustained throughput.
Paste this snippet into any page to show a live fit card.
<iframe src="https://willitrunai.com/embed/hf-unsloth--qwen3-5-122b-a10b-gguf-on-m3-ultra-96gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>
Preview: