QW

Qwen3.8 Flash Next

QwenGGUF by GuileUpdated 2026-08-29
Download on Hugging Face
Parameters177B
Context256K
Layers48
Architectureqwen4exp
Smallest quant0.6 GB
Quick answer

Qwen3.8 Flash Next needs about 120.9 GB of memory at Q4_K_M (a 119.6 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 29 quants, from 0.6 GB to 354.0 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 3.2 GB.

Qwen3.8 Flash Next – 177B parameter multimodal LLM

Qwen3.8 Flash Next is a 177B parameter model with a 262144 token context window. It supports text and image input and is built on the qwen4exp architecture. The model is available in multiple GGUF quantizations for llama.cpp.

Good at

  • Large context handling
  • Multimodal text and image
  • Efficient inference with imatrix

Limits

  • No speculative decoding
  • Requires llama.cpp b10665 or newer
  • License is other

Best for

  • Long‑form content generation
  • Multimodal chat applications
  • Research on large‑context LLMs

Who it’s for. Developers and researchers needing a high‑capacity multimodal LLM with a very long context window. The model is suitable for applications that can afford the memory and storage requirements of the larger quantizations.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—0.6 GB 1.9 GB
Fits Get file →
IQ1_S70.1 GB (2 parts) 71.4 GB
Too big Get file →
IQ1_M72.0 GB (2 parts) 73.3 GB
Too big Get file →
IQ2_XXS75.2 GB (2 parts) 76.5 GB
Too big Get file →
IQ2_XS77.7 GB (2 parts) 79.0 GB
Too big Get file →
IQ2_S77.8 GB (2 parts) 79.1 GB
Too big Get file →
IQ2_M80.4 GB (3 parts) 81.7 GB
Too big Get file →
Q2_K80.9 GB (3 parts) 82.2 GB
Too big Get file →
IQ3_XXS88.0 GB (3 parts) 89.3 GB
Too big Get file →
Q3_K_S89.4 GB (3 parts) 90.7 GB
Too big Get file →
IQ3_XS91.9 GB (3 parts) 93.3 GB
Too big Get file →
Q3_K_M92.0 GB (3 parts) 93.3 GB
Too big Get file →
Q3_K_L93.2 GB (3 parts) 94.5 GB
Too big Get file →
IQ3_M93.2 GB (3 parts) 94.5 GB
Too big Get file →
IQ4_XS97.7 GB (3 parts) 99.0 GB
Too big Get file →
IQ4_NL100.3 GB (3 parts) 101.6 GB
Too big Get file →
Q4_0100.6 GB (3 parts) 101.9 GB
Too big Get file →
Q2_K_L107.1 GB (4 parts) 108.5 GB
Too big Get file →
Q4_1111.3 GB (3 parts) 112.6 GB
Too big Get file →
Q4_K_S113.1 GB (3 parts) 114.4 GB
Too big Get file →
Q3_K_XL119.3 GB (4 parts) 120.7 GB
Too big Get file →
Q4_K_M119.6 GB (4 parts) 120.9 GB
Too big Get file →
Q5_K_S128.1 GB (4 parts) 129.4 GB
Too big Get file →
Q5_K_M134.7 GB (4 parts) 136.0 GB
Too big Get file →
Q4_K_L139.3 GB (5 parts) 140.6 GB
Too big Get file →
Q5_K_L151.1 GB (5 parts) 152.4 GB
Too big Get file →
Q6_K168.1 GB (5 parts) 169.4 GB
Too big Get file →
Q8_0188.3 GB (6 parts) 189.6 GB
Too big Get file →
BF16354.0 GB (9 parts) 355.3 GB
Too big Get file →

For 8‑GB VRAM, use IQ1_S (70.1 GB file, 71.4 GB memory at 8k context). For 16‑GB VRAM, IQ2_M (80.37 GB file, 81.7 GB memory) or Q2_K (80.93 GB file, 82.2 GB memory). For 24‑GB VRAM, Q4_K_M (119.6 GB file, 120.9 GB memory) offers a good balance of size and performance.

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

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 the main file (0.6 GB file, about 1.9 GB in use).
  • 6 GB VRAM: the best quant that fits is the main file (0.6 GB file, about 1.9 GB in use).
  • 8 GB VRAM: the best quant that fits is the main file (0.6 GB file, about 1.9 GB in use).
  • 12 GB VRAM: the best quant that fits is the main file (0.6 GB file, about 1.9 GB in use).
  • 16 GB VRAM: the best quant that fits is the main file (0.6 GB file, about 1.9 GB in use).
  • 24 GB VRAM: the best quant that fits is the main file (0.6 GB file, about 1.9 GB in use).
  • 32 GB VRAM: the best quant that fits is the main file (0.6 GB file, about 1.9 GB in use).
  • 48 GB VRAM: the best quant that fits is the main file (0.6 GB file, about 1.9 GB in use).

Which Qwen3.8 Flash Next quant should you choose?

Q4_K_M is the usual starting point for Qwen3.8 Flash Next: it balances file size and quality well. Here it is a 119.6 GB file.

The Q8_0 file is 188.3 GB, about 1.6 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.

If memory is tight, the smallest file is the first one in the table at 0.6 GB, with the quality trade-offs that come with fewer bits.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen4exp
Parameters177B (approximate, mixture of experts)
Layers48
Trained context262,144 tokens
KV cache (fp16)96 KB per token

Source

Quantized byGuile
FamilyQwen
Licenseother
Downloads (HF)1,296
Last updated2026-08-29

Frequently asked questions about Qwen3.8 Flash Next

How much VRAM does Qwen3.8 Flash Next need?

About 120.9 GB at Q4_K_M with an 8K context window. That is the 119.6 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 on an 8 GB GPU?

Yes, with the smaller files: the 0.6 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 GGUF?

It depends on the quant: file is 0.6 GB, IQ1_S is 70.1 GB, IQ1_M is 72.0 GB, IQ2_XXS is 75.2 GB, IQ2_XS is 77.7 GB, IQ2_S is 77.8 GB and 23 more.

What context length does Qwen3.8 Flash Next support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.8 GB at 8K and 3.2 GB at 32K.

Where can I download Qwen3.8 Flash Next GGUF?

The files are on Hugging Face in the Guile/Qwen3.8-Flash-Next-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 use?

Hugging Face lists the license as other. 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 on 8 GB VRAM?

The smallest quant, IQ1_S, requires about 71.4 GB memory at 8k context, so 8 GB VRAM is insufficient.

Does the model support image input?

Yes, image input is supported via the mmproj projector files included with each quant.

What llama.cpp release is needed?

All quants require llama.cpp release b10665 or newer to run.

Other GGUFs of Qwen3.8-Flash-Next

Similar models

Browse all →

More from Guile

All by Guile →

More Qwen models

All Qwen →

Other large models

All large models →