RunsOnMy ModelsGPUs & MacsBy VRAMBest GPUCan I run…?API pricesVRAM calculatorNew

Can Phi 3.5 Mini Instruct run on the Ryzen AI Max+ 395 128GB?

Yes. At 8K context the best fit is F16, using about 10.6 GB of the 109 GB usable on the Ryzen AI Max+ 395 128GB. Estimated speed ceiling: ~33 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctx128K ctxMax tok/s (est., 8K)
2-bit (Q2_K) •3.3 GB Fits4.8 GB Fits13.8 GB Fits49.8 GB Fits181
3-bit (Q3_K_M) •3.8 GB Fits5.3 GB Fits14.3 GB Fits50.3 GB Fits131
4-bit (Q4_K_M) •4.2 GB Fits5.7 GB Fits14.7 GB Fits50.7 GB Fits107
5-bit (Q5_K_M) •4.6 GB Fits6.1 GB Fits15.1 GB Fits51.1 GB Fits91
6-bit (Q6_K) •4.9 GB Fits6.4 GB Fits15.4 GB Fits51.4 GB Fits82
8-bit (Q8_0) •5.7 GB Fits7.2 GB Fits16.2 GB Fits52.2 GB Fits63
16-bit (F16/BF16)9.1 GB Fits10.6 GB Fits19.6 GB Fits55.6 GB Fits33

• real GGUF file size from bartowski/Phi-3.5-mini-instruct-GGUF. Speed is a bandwidth ceiling (256 GB/s), not a benchmark. Methodology.

What fits instead

"Runs" and "fits" mean the whole model in GPU memory at 8K context (4-bit or better for the model pick). Prices are US launch MSRPs, not today's street prices. "Popular" means the 60 most-downloaded models we track.

More

Phi 3.5 Mini Instruct on every GPU and Mac · Every model that runs on the Ryzen AI Max+ 395 128GB · Other models that fit 64 GB VRAM · Try other settings in the calculator

Guides: Best GPU for local LLMs by budget · Can I run gpt-oss, Llama 70B or Qwen3? · Mac vs PC for local LLMs

Data: Hugging Face Hub (microsoft/Phi-3.5-mini-instruct), AMD specs. Model data last checked 2026-11-17; refreshed hourly.