Devstral 2 123B Instruct needs ~76.9 GB VRAM. Mac Studio M3 Ultra 96GB has 69.1 GB. With Q3_K_S quantization, expect ~8 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
22.6 GB over capacity — needs offload or smaller quantization
Fit status
Too heavy
Decode
5.4 tok/s
TTFT
35565 ms
Safe context
4K
Memory
91.7 GB / 69.1 GB
Offload
20%
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 6.1 GB of extra host RAM just for the offloaded portion, before OS and other tools.
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | F | Too heavy | 5.6 tok/s | 18723 ms | 4K |
| Coding | F | Too heavy | 5.4 tok/s | 35565 ms | 4K |
| Agentic Coding | F | Too heavy | 5.1 tok/s | 55297 ms | 4K |
| Reasoning | F | Too heavy | 5.4 tok/s | 42031 ms | 4K |
| RAG | F | Too heavy | 5.1 tok/s | 69121 ms | 4K |
How Devstral 2 123B Instruct (123B 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 | 48.0 GB | Low | S91 |
Q3_K_S | 3 | 60.3 GB | Low | F0 |
Copy-paste commands to run Devstral 2 123B Instruct on your machine.
Run
lms load Devstral-2-123B-Instruct-2512 && lms server startUpgrade options
Makes the model fit on the accelerator instead of staying completely out of reach.
Removes host-memory offload, which is usually the single biggest latency and throughput win.
~$2,499 MSRP
Makes the model fit on the accelerator instead of staying completely out of reach.
Removes host-memory offload, which is usually the single biggest latency and throughput win.
~$3,999 MSRP
Makes the model fit on the accelerator instead of staying completely out of reach.
Removes host-memory offload, which is usually the single biggest latency and throughput win.
~$3,999 MSRP
Makes the model fit on the accelerator instead of staying completely out of reach.
Removes host-memory offload, which is usually the single biggest latency and throughput win.
~$12,000 MSRP
Yes, Mac Studio M3 Ultra 96GB can run Devstral 2 123B Instruct at Q3_K_S quantization (Very compromised (needs ~6.1 GB host RAM)). The recommended Q4_K_M requires 91.7 GB which exceeds available memory, but at Q3_K_S it needs only 76.9 GB. Expected decode speed: 7.8 tok/s.
Devstral 2 123B Instruct (123B parameters) requires approximately 91.7 GB at Q4_K_M quantization. On Mac Studio M3 Ultra 96GB, it fits at Q3_K_S using 76.9 GB.
The recommended quantization is Q4_K_M, but on Mac Studio M3 Ultra 96GB the best fitting quantization is Q3_K_S, which uses 76.9 GB.
On Mac Studio M3 Ultra 96GB, Devstral 2 123B Instruct achieves approximately 7.8 tokens per second decode speed with a time-to-first-token of 24665ms using Q3_K_S quantization.
For coding workloads, Devstral 2 123B Instruct on Mac Studio M3 Ultra 96GB receives a F grade with 5.4 tok/s and 4K context.
On Mac Studio M3 Ultra 96GB, Devstral 2 123B Instruct can safely use up to 4K tokens of context at Q3_K_S quantization. The model's official context limit is 256K, but available memory constrains the safe maximum.
Paste this snippet into any page to show a live fit card.
<iframe src="https://willitrunai.com/embed/devstral-2-123b-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:
| 4 |
68.9 GB |
| Medium |
| F0 |
Q4_K_M | 4 | 75.0 GB | Medium | F0 |
Q5_K_M | 5 | 88.6 GB | High | F0 |
Q6_K | 6 | 100.9 GB | High | F0 |
Q8_0 | 8 | 131.6 GB | Very High | F0 |
F16 | 16 | 252.2 GB | Maximum | F0 |
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.