llmat / text-generation updated 11 months ago

Qwen3-4B-Instruct-2507-NVFP4

Name: Qwen3-4B-Instruct-2507-NVFP4
Author: llmat

NVFP4-quantized version of Qwen/Qwen3-4B-Instruct-2507 produced with llmcompressor.

Params

2.8 B

Context

262,144

Downloads 30d

161 K

Likes

Commercial use: allowed apache-2.0 Not gated SAFETENSORS 1 languages View on Hugging Face ↗

Download history

daily snapshots · 10 days

214 K161 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	u8	3.6 GB	4.9 GB	✅ Runs comfortably
model.safetensors (bf16, full)	bf16 + 256K ctx	3.6 GB	18.0 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/qwen3-4b-instruct-2507-nvfp4
{
  "hf_id": "llmat/Qwen3-4B-Instruct-2507-NVFP4",
  "params_b": 2.82,
  "context_length": 262144,
  "license": { "id": "apache-2.0", "commercial": "yes" },
  "downloads_30d": 160914,
  "vram_estimates": [
    { "quant": "u8", "gb": 4.9 }
  ],
  "updated_at": "2026-07-28T18:07:54Z"
}

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

Specifications

Architecture: Qwen3ForCausalLM
Parameters: 2.8 B
Tensor type: U8
Context length: 262,144
Vocabulary: 151,936
Layers / heads: 36 / 32
Licence: apache-2.0
First seen on the Hub: 2025-08-27
Base model: Qwen3-4B-Instruct-2507
Training datasets: undisclosed
Added to our catalog: 2026-07-28

Family

Base model and the most-downloaded derivatives in the catalog.

Base Qwen3-4B-Instruct-2507 Qwen 3.2 M

Compare with

Qwen3-0.6B 29.7 M Qwen3-8B 16.3 M Qwen2.5-1.5B-Instruct 14.0 M Llama-3.2-1B-Instruct 10.4 M Qwen3-1.7B 7.8 M Qwen2.5-3B-Instruct 6.0 M

Appears in

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