Phi 3 mini 4k instruct

Phi 3 Mini 4K Instruct is a 3.82-billion-parameter dense transformer from Microsoft, trained on 4.9 trillion tokens with a focus on high-quality synthetic data for reasoning and code. It delivers performance competitive with much larger 7B models on math and coding benchmarks while keeping a minimal memory footprint. The model supports English and French, with capabilities in code generation and instruction following. A 4K context window and flash attention make it well suited for edge and resource-constrained deployments, and it quantizes well to GGUF for local inference.

Hardware Configuration

Vendor Select vendor...

Product —

Platform Select platform...

Family —

Model —

VRAM —

System RAM (GB) Optional — for precise deployment recommendations

Clear Configuration

Quantization	Quality	Size	Fit
FP16	Full precision	7.12 GB	—
Q4	Low	2.23 GB	—

Last updated: March 5, 2026