QW

Qwen3.8-Flash-Next-DFlash2

QwenGGUF by incoaiUpdated 2026-10-20
Download on Hugging Face
Parameters603M
Context256K
Layers5
Architecturedflash
Smallest quant0.4 GB
Quick answer

Qwen3.8-Flash-Next-DFlash2 needs about 1.0 GB of memory at Q4_K_M (a 0.4 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 5 quants, from 0.4 GB (Q4_K_M) to 1.2 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 0.7 GB.

Qwen3.8-Flash-Next-DFlash2: a draft model for speculative decoding

Qwen3.8-Flash-Next-DFlash2 is a draft model built on the Qwen3.8-Flash-Next architecture. It is designed for speculative decoding, predicting whole blocks of tokens and selecting a coherent path. The model is available in GGUF format and is licensed under apache-2.0.

Good at

  • speculative decoding
  • block‑level token prediction
  • drafting for target models

Limits

  • no standalone inference support
  • requires a target model for verification
  • no vision capability

Best for

  • enhancing inference speed in large‑scale deployments
  • drafting tokens for Qwen3.8-Flash-Next
  • optimizing speculative decoding pipelines

Who it’s for. Developers building speculative decoding systems that need a lightweight draft model. The model is suitable for integration with engines that support DFlash 2, such as SGLang, and for users who want to reduce inference latency by drafting tokens before verification.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_M0.4 GB 1.0 GB
Fits Get file →
Q5_K_M0.4 GB 1.1 GB
Fits Get file →
Q6_K0.5 GB 1.2 GB
Fits Get file →
Q8_00.7 GB 1.3 GB
Fits Get file →
BF161.2 GB 1.9 GB
Fits Get file →

For 8 GB VRAM, the Q4_K_M quantization (0.36 GB file, ~1 GB memory at 8 k context) is recommended. If more memory is available, Q5_K_M (0.43 GB, ~1.1 GB memory) or Q6_K (0.51 GB, ~1.2 GB memory) can be used. The BF16 quantization (1.22 GB, ~1.9 GB memory) requires larger VRAM and is suitable for high‑performance setups.

What GPU do you need for Qwen3.8-Flash-Next-DFlash2?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: the best quant that fits is BF16 (1.2 GB file, about 1.9 GB in use).

Which Qwen3.8-Flash-Next-DFlash2 quant should you choose?

Q4_K_M is the usual starting point for Qwen3.8-Flash-Next-DFlash2: it balances file size and quality well. Here it is a 0.4 GB file.

The Q8_0 file is 0.7 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturedflash
Parameters603M
Layers5
Trained context262,144 tokens
KV cache (fp16)20 KB per token

Source

Quantized byincoai
FamilyQwen
Licenseapache-2.0
Downloads (HF)3,949
Last updated2026-10-20

Frequently asked questions about Qwen3.8-Flash-Next-DFlash2

How much VRAM does Qwen3.8-Flash-Next-DFlash2 need?

About 1.0 GB at Q4_K_M with an 8K context window. That is the 0.4 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run Qwen3.8-Flash-Next-DFlash2 on an 8 GB GPU?

Yes, with BF16: the 1.2 GB file needs about 1.9 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Qwen3.8-Flash-Next-DFlash2 GGUF?

It depends on the quant: Q4_K_M is 0.4 GB, Q5_K_M is 0.4 GB, Q6_K is 0.5 GB, Q8_0 is 0.7 GB, BF16 is 1.2 GB.

What context length does Qwen3.8-Flash-Next-DFlash2 support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.2 GB at 8K and 0.7 GB at 32K.

Where can I download Qwen3.8-Flash-Next-DFlash2 GGUF?

The files are on Hugging Face in the incoai/Qwen3.8-Flash-Next-DFlash2-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does Qwen3.8-Flash-Next-DFlash2 use?

Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run Qwen3.8-Flash-Next-DFlash2 on 8 GB VRAM?

Yes, the Q4_K_M quantization requires about 1 GB of VRAM at 8 k context, which fits within an 8 GB budget.

Does this model support vision tasks?

No, the model has no vision capability.

What license applies to this model?

The model is licensed under apache-2.0.

Other GGUFs of Qwen3.8-Flash-Next

Similar models

Browse all →

More from incoai

All by incoai →

More Qwen models

All Qwen →

Other tiny models

All tiny models →