GL

Glm53 Dflash2 Draft

GLMGGUF by AnbeeldUpdated 2026-08-29
Download on Hugging Face
Parameters2.46B
Context1024K
Layers6
Architecturedflash
Smallest quant0.9 GB
Quick answer

Glm53 Dflash2 Draft needs about 2.1 GB of memory at Q4_K_M (a 1.4 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 7 quants, from 0.9 GB (Q2_K) to 4.9 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 0.8 GB.

GLM-5.3 DFlash2 Draft – a speculative decoding draft model

GLM-5.3 DFlash2 Draft is a block‑diffusion drafter for speculative decoding. It runs inside a speculative decoding server and drafts tokens for the target model to verify. The model is based on GLM‑5.3 and uses the dflash architecture.

Good at

  • Speculative decoding
  • Drafting token blocks
  • Supporting target model verification

Limits

  • No vision support
  • Requires a target model
  • License is cc‑by‑nc‑nd‑4.0

Best for

  • Speeding up inference with speculative decoding
  • Drafting tokens for GLM‑5.3
  • Testing draft quality in a server environment

Who it’s for. Developers building speculative decoding pipelines and researchers testing draft models. It is useful for those who need a lightweight drafter that works with GLM‑5.3.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K0.9 GB 1.6 GB
Fits Get file →
Q3_K_M1.2 GB 1.9 GB
Fits Get file →
Q4_K_M1.4 GB 2.1 GB
Fits Get file →
Q5_K_M1.7 GB 2.4 GB
Fits Get file →
Q6_K2.0 GB 2.7 GB
Fits Get file →
Q8_02.6 GB 3.3 GB
Fits Get file →
BF164.9 GB 5.6 GB
Fits Get file →

For 8 GB VRAM, use Q4_K_M (file size 1.44 GB, estimated memory 2.1 GB at 8 k context). For 16 GB VRAM, Q6_K (2.03 GB, 2.7 GB memory) or Q8_0 (2.62 GB, 3.3 GB memory) are suitable. Higher precision BF16 requires 5.6 GB memory and 4.93 GB file size.

What GPU do you need for Glm53 Dflash2 Draft?

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 (2.6 GB file, about 3.3 GB in use).
  • 6 GB VRAM: the best quant that fits is BF16 (4.9 GB file, about 5.6 GB in use).

Which Glm53 Dflash2 Draft quant should you choose?

Q4_K_M is the usual starting point for Glm53 Dflash2 Draft: it balances file size and quality well. Here it is a 1.4 GB file.

The Q8_0 file is 2.6 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.

If memory is tight, the smallest file is Q2_K at 0.9 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturedflash
Parameters2.46B
Layers6
Trained context1,048,576 tokens
KV cache (fp16)24 KB per token

Source

Quantized byAnbeeld
FamilyGLM
Base modelzai-org/GLM-5.3
Downloads (HF)627
Last updated2026-08-29

Frequently asked questions about Glm53 Dflash2 Draft

How much VRAM does Glm53 Dflash2 Draft need?

About 2.1 GB at Q4_K_M 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 Glm53 Dflash2 Draft on an 8 GB GPU?

Yes, with BF16: the 4.9 GB file needs about 5.6 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Glm53 Dflash2 Draft GGUF?

It depends on the quant: Q2_K is 0.9 GB, Q3_K_M is 1.2 GB, Q4_K_M is 1.4 GB, Q5_K_M is 1.7 GB, Q6_K is 2.0 GB, Q8_0 is 2.6 GB and 1 more.

What context length does Glm53 Dflash2 Draft support?

The GGUF header lists a trained context length of 1,048,576 tokens (1024K). 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 Glm53 Dflash2 Draft GGUF?

The files are on Hugging Face in the Anbeeld/GLM-5.3-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 Glm53 Dflash2 Draft use?

Hugging Face lists the license as cc-by-nc-nd-4.0. 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 GLM-5.3 DFlash2 Draft on 8 GB VRAM?

Yes, the Q4_K_M quantization uses about 2.1 GB memory at 8 k context, fitting within 8 GB VRAM.

Does this model support vision tasks?

No, GLM‑5.3 DFlash2 Draft does not include vision capabilities.

What license applies to this model?

The model is released under cc‑by‑nc‑nd‑4.0, allowing non‑commercial use with attribution and no derivatives.

Other GGUFs of GLM-5.3

Similar models

Browse all →

More from Anbeeld

All by Anbeeld →

More GLM models

All GLM →

Other tiny models

All tiny models →