 |
VOOZH | about |
Mistral Small 24B Instruct 2501 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 24B Instruct 2501 (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 24B Instruct 2501 on your machine.
Run
lms load hf-maziyarpanahi--mistral-small-24b-instruct-2501-gguf && 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.
~$599 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.
~$1,999 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.
~$15,000 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.
~$15,000 MSRP
| 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 24B Instruct 2501 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.
