microsoft / text-generation updated 7 months ago

Phi-mini-MoE-instruct

Name: Phi-mini-MoE-instruct
Author: microsoft

Phi-mini-MoE is a lightweight Mixture of Experts (MoE) model with 7.6B total parameters and 2.4B activated parameters. It is compressed and distilled from the base model shared by Phi-3.5-MoE and GRIN-MoE using the SlimMoE approach, then post-trained via supervised fine-tuning and direct preference optimization for instruction following and safety. The model is trained on Phi-3 synthetic data and filtered public docu...

Params

7.7 B

Context

4,096

Downloads 30d

189 K

Likes

Commercial use: allowed mit Not gated SAFETENSORS 1 languages View on Hugging Face ↗

Download history

daily snapshots · 10 days

189 K178 K

Jul 28Jul 31Aug 3Aug 6

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

File	Quant	Size	Est. VRAM	Verdict on RTX 4090 · 24 GB
model.safetensors	bf16	15.3 GB	18.5 GB	✅ Runs comfortably

Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Run it

copy-paste, exact tags checked against the Hub

~ · curl · api/v1

$ curl -s https://aimodelscomparison.com/api/v1/models/phi-mini-moe-instruct
{
  "hf_id": "microsoft/Phi-mini-MoE-instruct",
  "params_b": 7.65,
  "context_length": 4096,
  "license": { "id": "mit", "commercial": "yes" },
  "downloads_30d": 189297,
  "vram_estimates": [
    { "quant": "bf16", "gb": 18.5 }
  ],
  "updated_at": "2026-07-28T18:08:29Z"
}

est. VRAM —on RTX 4090 · 24 GBJSON API →

Specifications

Architecture: PhiMoEForCausalLM
Parameters: 7.7 B
Tensor type: BF16
Context length: 4,096
Vocabulary: 32,064
Layers / heads: 32 / 32
Licence: mit
First seen on the Hub: 2025-06-23
Training datasets: undisclosed
Added to our catalog: 2026-07-28

Compare with

Qwen3-8B 16.3 M Qwen2.5-7B-Instruct 12.1 M Qwen3.6-35B-A3B-NVFP4 11.3 M Llama-3.1-8B-Instruct 8.0 M Qwen2.5-7B-Instruct-AWQ 4.6 M granite-4.1-8b 4.2 M

Appears in

Best open coding models Commercial-use text models Models that fit in 12 GB