Mistral Small 3.2 24B Instruct 2506 needs ~18.8 GB VRAM. Intel Arc A770 16GB has 16.0 GB. With NVFP4 quantization, expect ~11 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
4.0 GB over capacity — needs offload or smaller quantization
Fit status
Too heavy
Decode
8.1 tok/s
TTFT
23867 ms
Safe context
4K
Memory
20.0 GB / 16.0 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.
Runtime ecosystem is narrower than CUDA
Intel GPUs can look attractive on memory per dollar, but local AI tooling, kernels, and model coverage are still broader and easier on CUDA today.
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.
Prefer CUDA if you want the path of least resistance
If your goal is maximum runtime coverage, easier troubleshooting, and better support for new local AI releases, CUDA is usually still the safer upgrade path.
Buy headroom, not only minimum fit
A slightly larger memory tier gives you safer context growth and makes the recommendation more future-proof.
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | D | Very compromised (needs ~2 GB host RAM) | 9.5 tok/s | 11162 ms | 4K |
| Coding | F | Too heavy | 8.1 tok/s | 23867 ms | 4K |
| Agentic Coding | F | Too heavy | 6.1 tok/s | 45823 ms | 4K |
| Reasoning | F | Too heavy | 8.1 tok/s | 28207 ms | 4K |
| RAG | F | Too heavy | 6.1 tok/s | 57279 ms | 4K |
How Mistral Small 3.2 24B Instruct 2506 (24B params) fits at each quantization level on Intel Arc A770 16GB (16.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 9.4 GB | Low | C51 |
Q3_K_SBest for your GPU | 3 | 11.8 GB | Low | C51 |
Copy-paste commands to run Mistral Small 3.2 24B Instruct 2506 on your machine.
Run
lms load hf-unsloth--mistral-small-3-2-24b-instruct-2506-gguf && lms server startUpgrade options
Intel Arc Pro B60 24GBBudget pickMakes 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.
~$599 MSRP
RTX 5090 32GBBiggest leapMakes 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.
~$1,999 MSRP
Intel Data Center GPU Max 1550 128GBBest valueMakes 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.
~$15,000 MSRP
Gaudi 3 128GBIntel upgradeMakes 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.
~$15,000 MSRP
Yes, Intel Arc A770 16GB can run Mistral Small 3.2 24B Instruct 2506 at NVFP4 quantization (Very compromised (needs ~2 GB host RAM)). The recommended Q4_K_M requires 20.0 GB which exceeds available memory, but at NVFP4 it needs only 18.8 GB. Expected decode speed: 10.6 tok/s.
Mistral Small 3.2 24B Instruct 2506 (24B parameters) requires approximately 20.0 GB at Q4_K_M quantization. On Intel Arc A770 16GB, it fits at NVFP4 using 18.8 GB.
The recommended quantization is Q4_K_M, but on Intel Arc A770 16GB the best fitting quantization is NVFP4, which uses 18.8 GB.
On Intel Arc A770 16GB, Mistral Small 3.2 24B Instruct 2506 achieves approximately 10.6 tokens per second decode speed with a time-to-first-token of 18313ms using NVFP4 quantization.
For coding workloads, Mistral Small 3.2 24B Instruct 2506 on Intel Arc A770 16GB receives a F grade with 8.1 tok/s and 4K context.
Paste this snippet into any page to show a live fit card.
<iframe src="https://willitrunai.com/embed/hf-unsloth--mistral-small-3-2-24b-instruct-2506-gguf-on-arc-a770-16gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>
Preview:
| 4 |
13.4 GB |
| Medium |
| F0 |
Q4_K_M | 4 | 14.6 GB | Medium | F0 |
Q5_K_M | 5 | 17.3 GB | High | F0 |
Q6_K | 6 | 19.7 GB | High | F0 |
Q8_0 | 8 | 25.7 GB | Very High | F0 |
F16 | 16 | 49.2 GB | Maximum | F0 |
On Intel Arc A770 16GB, Mistral Small 3.2 24B Instruct 2506 can safely use up to 4K tokens of context at NVFP4 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.
Often yes, if your goal is the easiest setup and the widest runtime support. Intel can offer attractive memory capacity, but CUDA still tends to win on tooling maturity, guides, kernels, and model coverage for local AI.