sakamakismile / text-generation updated 3 months ago

Qwen3.6-27B-Text-NVFP4-MTP

Name: Qwen3.6-27B-Text-NVFP4-MTP
Author: sakamakismile

NVFP4-quantized text-only sibling of Qwen/Qwen3.6-27B, with the MTP (Multi-Token Prediction) head restored in bf16 so speculative decoding actually works.

Params

16.7 B

Context

—

Downloads 30d

747 K

Likes

Commercial use: allowed apache-2.0 Not gated SAFETENSORS 11 languages View on Hugging Face ↗

Download history

daily snapshots · 10 days

747 K545 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	19.6 GB	24.6 GB	❌ Won’t fit

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/qwen3-6-27b-text-nvfp4-mtp
{
  "hf_id": "sakamakismile/Qwen3.6-27B-Text-NVFP4-MTP",
  "params_b": 16.67,
  "context_length": null,
  "license": { "id": "apache-2.0", "commercial": "yes" },
  "downloads_30d": 747066,
  "vram_estimates": [
    { "quant": "u8", "gb": 24.6 }
  ],
  "updated_at": "2026-07-28T18:04:55Z"
}

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

Specifications

Architecture: Qwen3_5ForConditionalGeneration
Parameters: 16.7 B
Tensor type: U8
Licence: apache-2.0
First seen on the Hub: 2026-04-25
Base model: Qwen3.6-27B
Training datasets: undisclosed
Added to our catalog: 2026-07-28

Family

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

Base Qwen3.6-27B Qwen 7.0 M

Compare with

Qwen3-8B 16.3 M Qwen2.5-7B-Instruct 12.1 M Qwen3.6-35B-A3B-NVFP4 11.3 M gpt-oss-20b 8.6 M Qwen3-32B 8.3 M Llama-3.1-8B-Instruct 8.0 M

Appears in

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