GLM-5.3-Flash-DFlash2

Blog | GitHub

This repository contains the DFlash 2 draft model for zai-org/GLM-5.3-Flash. It is not a standalone language model: it runs inside a speculative decoding server and drafts tokens for the target model to verify.

DFlash 2 is a block-diffusion drafter for speculative decoding. It predicts a whole block of tokens in a single pass and keeps the top candidates at every position. A lightweight selector then traces one coherent path through them. Two-tap dynamic convolutions in the backbone keep the draft from decaying toward the end of the block. Decoding is lossless: greedy output matches the target model exactly, and sampling preserves its distribution.

Quick Start

Serve with SGLang:

pip install "sglang[all] @ git+https://github.com/sgl-project/sglang.git@refs/pull/36708/head#subdirectory=python"

sglang serve \
  --model-path zai-org/GLM-5.3-Flash \
  --trust-remote-code \
  --speculative-algorithm DFLASH \
  --speculative-draft-model-path incoai/GLM-5.3-Flash-DFlash2 \
  --speculative-draft-attention-backend fa4

See the blog post for more details.

Evaluation

Benchmark numbers for this checkpoint are being re-measured and will be published here once final.

Citation

If you find DFlash 2 useful, please cite:

@misc{inco2026dflash2,
  title  = {{DFlash 2: Keep Drafting Parallel}},
  author = {{Inco AI}},
  year   = {2026},
  month  = {August},
  url    = {https://inco.ai/blog/dflash2/}
}

Please also cite the original DFlash paper:

@inproceedings{chen2026dflash,
  title     = {{DFlash: Block Diffusion for Flash Speculative Decoding}},
  author    = {Chen, Jian and Liang, Yesheng and Liu, Zhijian},
  booktitle = {International Conference on Machine Learning (ICML)},
  year      = {2026}
}