microsoft / text-generation Open weights updated 11 months ago

phi-2

Name: phi-2
Author: microsoft

Phi-2 is a Transformer with 2.7 billion parameters. It was trained using the same data sources as Phi-1.5, augmented with a new data source that consists of various NLP synthetic texts and filtered websites (for safety and educational value). When assessed against benchmarks testing common sense, language understanding, and logical reasoning, Phi-2 showcased a nearly state-of-the-art performance among models with les...

Params

2.8 B

Context

2,048

Downloads 30d

178 K

Likes

3,534

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

Download history

daily snapshots · 106 days

▲ 408 K in the last 30 days (69.7%)

1.6 M177 K

Jul 28Sep 1Oct 6Nov 10

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	f16	5.6 GB	7.0 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-2
{
  "hf_id": "microsoft/phi-2",
  "params_b": 2.78,
  "context_length": 2048,
  "license": { "id": "mit", "commercial": "yes" },
  "downloads_30d": 177546,
  "vram_estimates": [
    { "quant": "f16", "gb": 7.0 }
  ],
  "updated_at": "2026-07-28T18:03:48Z"
}

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

Specifications

Architecture: PhiForCausalLM
Parameters: 2.8 B
Tensor type: F16
Context length: 2,048
Vocabulary: 51,200
Layers / heads: 32 / 32
Licence: mit
First seen on the Hub: 2023-12-13
Training datasets: undisclosed
Added to our catalog: 2026-07-28

Compare with any text-generation model

Popular comparisons

Qwen3-0.6B 20.7 M Bonsai-27B-mlx-1bit 1.0 M SmolLM3-3B-Base 213 K granite-4.1-8b 192 K PowerMoE-3b 171 K Qwen2-1.5B-Instruct 144 K