unsloth / image-text-to-text updated 2 months ago

Qwen3.6-35B-A3B-NVFP4

Name: Qwen3.6-35B-A3B-NVFP4
Author: unsloth

1.56x faster throughput than other NVFP4 quants. This is an Unsloth NVFP4 quantized checkpoint calibrated on a mixture of our Unsloth dataset + UltraChat dataset. Works on a 32GB VRAM GPU. Benchmarks on 1xB200 128 concurrency. Use the 35B NVFP4 Fast version for 1.79x faster at a little less accuracy

Params

24.6 B

Context

—

Downloads 30d

1.3 M

Likes

118

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

Download history

daily snapshots · 56 days

▲ 1.2 M in the last 30 days (47.6%)

2.6 M923 K

Jul 28Aug 15Sep 3Sep 21

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	26.5 GB	33.3 GB	❌ Won’t fit

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

Specifications

Architecture: Qwen3_5MoeForConditionalGeneration
Parameters: 24.6 B
Tensor type: U8
Licence: apache-2.0
First seen on the Hub: 2026-04-23
Base model: Qwen3.6-35B-A3B
Training datasets: undisclosed
Added to our catalog: 2026-07-28

Family

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

Base Qwen3.6-35B-A3B Qwen 3.3 M

Compare with any image-text-to-text model

Popular comparisons

Qwen3-VL-8B-Instruct 19.7 M gemma-4-26B-A4B-it 10.0 M Qwen3.6-35B-A3B-FP8 9.5 M gemma-4-31B-it 9.0 M Qwen3.8-27B 7.3 M Qwen3.8-27B-FP8 6.9 M