Home / Models / Phi-3.5 Mini Instruct
Phi-3.5 Mini Instruct
3.8B parameters · native context 131,072 · phi3
Estimated VRAM requirement
Weight memory uses parameter count × bits per weight, plus a quantization overhead factor. KV cache uses standard GQA formulas. Values are calculated estimates, not measured allocations.
| Quantization | Estimated weights | 2K total | 4K total | 8K total | 16K total | 32K total | 64K total | 128K total |
|---|---|---|---|---|---|---|---|---|
| FP16 | 7.3 GB | 9.3 GB | 10.1 GB | 11.6 GB | 14.6 GB | 20.6 GB | 32.6 GB | 56.6 GB |
| Q8 | 4.1 GB | 5.9 GB | 6.7 GB | 8.2 GB | 11.2 GB | 17.2 GB | 29.2 GB | 53.2 GB |
| Q6 | 3.2 GB | 4.9 GB | 5.7 GB | 7.2 GB | 10.2 GB | 16.2 GB | 28.2 GB | 52.2 GB |
| Q5 | 2.7 GB | 4.4 GB | 5.2 GB | 6.7 GB | 9.7 GB | 15.7 GB | 27.7 GB | 51.7 GB |
| Q4 | 2.2 GB | 3.9 GB | 4.7 GB | 6.2 GB | 9.2 GB | 15.2 GB | 27.2 GB | 51.2 GB |
| Q3 | 1.8 GB | 3.4 GB | 4.2 GB | 5.7 GB | 8.7 GB | 14.7 GB | 26.7 GB | 50.7 GB |
Find a GPU for this model
Run at the quantization and context selected below. Defaults are Q4 and 8K when you do not change the controls. Results are full-VRAM fits only — not a speed ranking.
- Run at
- Q4 · 8K context
- Required VRAM
- 6.2 GB
Lowest-cost full-VRAM option
Sorted by lowest current Amazon price among GPUs that fit entirely in VRAM. This is not a performance ranking.
NVIDIA Quadro P4000 8 GB
- GPU VRAM
- 8 GB advertised
- Required VRAM
- 6.2 GB
- Usable GPU VRAM
- 7.2 GB
- VRAM headroom
- 1.0 GB — leftover usable memory after the calculated requirement, not a speed ranking
- Compatible
- Yes — fits entirely in VRAM
- Product SKUs
- 6 mapped · 1 Amazon-matched
- Amazon
- $216.44 current Amazon price
NVIDIA Tesla K40 12 GB
- GPU VRAM
- 12 GB advertised
- Required VRAM
- 6.2 GB
- Usable GPU VRAM
- 10.8 GB
- VRAM headroom
- 4.6 GB — leftover usable memory after the calculated requirement, not a speed ranking
- Compatible
- Yes — fits entirely in VRAM
- Product SKUs
- 2 mapped · 1 Amazon-matched
- Amazon
- $262.48 current Amazon price
NVIDIA RTX 3060 Ti 8 GB
- GPU VRAM
- 8 GB advertised
- Required VRAM
- 6.2 GB
- Usable GPU VRAM
- 7.2 GB
- VRAM headroom
- 1.0 GB — leftover usable memory after the calculated requirement, not a speed ranking
- Compatible
- Yes — fits entirely in VRAM
- Product SKUs
- 32 mapped · 2 Amazon-matched
- Amazon
- $283.22 current Amazon price
More VRAM headroom
Sorted by leftover usable VRAM after the calculated requirement. Headroom is not tokens/sec.
NVIDIA RTX 6000 48 GB
- GPU VRAM
- 48 GB advertised
- Required VRAM
- 6.2 GB
- Usable GPU VRAM
- 43.2 GB
- VRAM headroom
- 37.0 GB — leftover usable memory after the calculated requirement, not a speed ranking
- Compatible
- Yes — fits entirely in VRAM
- Product SKUs
- 1 mapped · 0 Amazon-matched
- Amazon
- Check price on Amazon
NVIDIA RTX 8000 48 GB
- GPU VRAM
- 48 GB advertised
- Required VRAM
- 6.2 GB
- Usable GPU VRAM
- 43.2 GB
- VRAM headroom
- 37.0 GB — leftover usable memory after the calculated requirement, not a speed ranking
- Compatible
- Yes — fits entirely in VRAM
- Product SKUs
- 1 mapped · 0 Amazon-matched
- Amazon
- Check availability
NVIDIA RTX A6000 48 GB
- GPU VRAM
- 48 GB advertised
- Required VRAM
- 6.2 GB
- Usable GPU VRAM
- 43.2 GB
- VRAM headroom
- 37.0 GB — leftover usable memory after the calculated requirement, not a speed ranking
- Compatible
- Yes — fits entirely in VRAM
- Product SKUs
- 2 mapped · 0 Amazon-matched
- Amazon
- Check availability
Other compatible GPUs
AMD Radeon RX 470 8 GB
- GPU VRAM
- 8 GB advertised
- Required VRAM
- 6.2 GB
- Usable GPU VRAM
- 7.2 GB
- VRAM headroom
- 1.0 GB — leftover usable memory after the calculated requirement, not a speed ranking
- Compatible
- Yes — fits entirely in VRAM
- Product SKUs
- 1 mapped · 0 Amazon-matched
- Amazon
- Check availability
AMD Radeon RX 480 8 GB
- GPU VRAM
- 8 GB advertised
- Required VRAM
- 6.2 GB
- Usable GPU VRAM
- 7.2 GB
- VRAM headroom
- 1.0 GB — leftover usable memory after the calculated requirement, not a speed ranking
- Compatible
- Yes — fits entirely in VRAM
- Product SKUs
- 4 mapped · 0 Amazon-matched
- Amazon
- Check availability
AMD Radeon RX 5500 XT 8 GB
- GPU VRAM
- 8 GB advertised
- Required VRAM
- 6.2 GB
- Usable GPU VRAM
- 7.2 GB
- VRAM headroom
- 1.0 GB — leftover usable memory after the calculated requirement, not a speed ranking
- Compatible
- Yes — fits entirely in VRAM
- Product SKUs
- 4 mapped · 0 Amazon-matched
- Amazon
- Check price on Amazon
Compatible GPUs
Status is a calculated memory fit against usable VRAM (advertised × 0.9). Open a GPU to see product SKUs and Amazon listings.
Sources
Model specifications: official repository microsoft/Phi-3.5-mini-instruct · revision main. Compatibility: RigForAI calculator v1.0.0 (estimated VRAM). GPU products: Icecat. Amazon is a separate affiliate match and is not used as a spec source.