Atomic Chat / Qwen3.5 9B DFlash needs about 2.1 GB of memory at Q8_0 (a 1.4 GB file) with an 8K context window, so it fits on a 4 GB GPU. Longer context needs more: at 32K tokens the KV cache alone adds about 0.8 GB.
Atomic Chat / Qwen3.5 9B DFlash – a 1.3B parameter draft model with 256K context
Atomic Chat has self‑quantized Qwen3.5 9B DFlash into GGUF. The model contains 1.3 billion parameters and supports a 262,144‑token context length. It is designed for speculative decoding with DFlash and runs fully offline.
Good at
- High‑throughput speculative decoding
- Long‑context generation up to 256K tokens
- Low‑latency inference with Q8_0 quant
Limits
- No vision or multimodal support
- Requires DFlash‑enabled server
- License is Apache‑2.0
Best for
- Chatbots needing long context
- Batch inference on single GPU
- Research on draft‑based decoding
Who it’s for. Developers and researchers who need a lightweight, long‑context LLM for offline use. The model is ideal for environments where GPU memory is limited but high throughput is required.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q8_0 | 1.4 GB | 2.1 GB | Fits | Get file → |
The only available quant is Q8_0, 1.38 GB. It requires roughly 2.1 GB VRAM at 8 k context. For 8 GB VRAM, Q8_0 fits comfortably with room for context buffers. No lower‑bit quants are provided, so Q8_0 is the recommended choice.
What GPU do you need for Atomic Chat / Qwen3.5 9B DFlash?
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 Q8_0 (1.4 GB file, about 2.1 GB in use).
Which Atomic Chat / Qwen3.5 9B DFlash quant should you choose?
Q8_0 (1.4 GB) is the middle option in this repository. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Atomic Chat / Qwen3.5 9B DFlash
How much VRAM does Atomic Chat / Qwen3.5 9B DFlash need?
About 2.1 GB at Q8_0 with an 8K context window. That is the 1.4 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 Atomic Chat / Qwen3.5 9B DFlash on an 8 GB GPU?
Yes, with Q8_0: the 1.4 GB file needs about 2.1 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Atomic Chat / Qwen3.5 9B DFlash GGUF?
It depends on the quant: Q8_0 is 1.4 GB.
What context length does Atomic Chat / Qwen3.5 9B DFlash 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 0.8 GB at 32K.
Where can I download Atomic Chat / Qwen3.5 9B DFlash GGUF?
The files are on Hugging Face in the AtomicChat/Qwen3.5-9B-DFlash-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 Atomic Chat / Qwen3.5 9B DFlash 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 Qwen3.5 9B DFlash on 8 GB VRAM?
Yes, the Q8_0 quant uses about 2.1 GB VRAM at 8 k context, leaving space for buffers.
Does this model support image input?
No, the model has no vision or multimodal capabilities.
What license applies to this GGUF?
The GGUF is released under Apache‑2.0, the same license as the original model.