Total runs:
64.5M
Run Growth:
-9.0M
Growth Rate:
-13.93%
Sponsored by
APIMart
BAAI / Emu3.5-VisionTokenizer
Model's Last Updated: January 04 2026
Introduction of Emu3.5-VisionTokenizer
Model Details of Emu3.5-VisionTokenizer
Emu3.5: Native Multimodal Models are World Learners
Emu3.5 Team, BAAI
| 🔹 | Core Concept | Description |
|---|---|---|
| 🧠 | Unified World Modeling | Predicts the next state jointly across vision and language , enabling coherent world modeling and generation . |
| 🧩 | End-to-End Pretraining | Trained with a unified next-token prediction objective over interleaved vision–language sequences . |
| 📚 | Over 10T+ Multimodal Tokens | Pre-trained on over 10 trillion interleaved tokens from video frames and transcripts , capturing spatiotemporal structure . |
| 🔄 | Native Multimodal I/O | Processes and generates interleaved visual–text sequences without modality adapters or task-specific heads . |
| 🎯 | RL Post-Training | Large-scale reinforcement learning enhances reasoning , compositionality , and generation quality . |
| ⚡ | Discrete Diffusion Adaptation (DiDA) | Converts sequential decoding → bidirectional parallel prediction , achieving ≈20× faster inference without performance loss . |
| 🖼️ | Versatile Generation | Excels in long-horizon vision–language generation , any-to-image (X2I) synthesis, and text-rich image creation . |
| 🌐 | Generalizable World Modeling | Enables spatiotemporally consistent world exploration , and open-world embodied manipulation across diverse scenarios. |
| 🏆 | Performance Benchmark | Matches Gemini 2.5 Flash Image (Nano Banana) on image generation/editing , and outperforms on interleaved generation tasks . |
Table of Contents
1. Model & Weights
2. Quick Start
Environment Setup
git clone https://github.com/baaivision/Emu3.5
cd Emu3.5
pip install -r requirements.txt
pip install flash_attn==2.8.3 --no-build-isolation
Configuration
Edit
configs/config.py
to set:
-
Paths:
model_path,vq_path -
Task template:
task_type in {t2i, x2i, howto, story, explore, vla},use_imagecontrols<|IMAGE|>usage (set to true when reference images are provided) -
Sampling:
sampling_params(classifier_free_guidance, temperature, top_k/top_p, etc.)
Run Inference
python inference.py --cfg configs/config.py
Protobuf outputs are written to
outputs/<exp_name>/proto/
. For better throughput, we recommend ≥2 GPUs.
Visualize Protobuf Outputs
To visualize generated protobuf files:
python src/utils/vis_proto.py --input <input_proto_file> --output <output_dir>
3. Schedule
- Inference Code
- Advanced Image Decoder
- Discrete Diffusion Adaptation(DiDA)
4. Citation
@misc{cui2025emu35nativemultimodalmodels,
title={Emu3.5: Native Multimodal Models are World Learners},
author={Yufeng Cui and Honghao Chen and Haoge Deng and Xu Huang and Xinghang Li and Jirong Liu and Yang Liu and Zhuoyan Luo and Jinsheng Wang and Wenxuan Wang and Yueze Wang and Chengyuan Wang and Fan Zhang and Yingli Zhao and Ting Pan and Xianduo Li and Zecheng Hao and Wenxuan Ma and Zhuo Chen and Yulong Ao and Tiejun Huang and Zhongyuan Wang and Xinlong Wang},
year={2025},
eprint={2510.26583},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2510.26583},
}
Runs of BAAI Emu3.5-VisionTokenizer on huggingface.co
1.2K
Total runs
0
24-hour runs
28
3-day runs
85
7-day runs
-685
30-day runs
More Information About Emu3.5-VisionTokenizer huggingface.co Model
Url of Emu3.5-VisionTokenizer
Emu3.5-VisionTokenizer huggingface.co Url
Provider of Emu3.5-VisionTokenizer huggingface.co
Other API from BAAI
Updated:February 22 2024
Total runs:
17.7M
Run Growth:
-918.6K
Growth Rate:
-5.15%
Total runs:
11.6M
Run Growth:
-1.1M
Growth Rate:
-9.17%
Total runs:
10.5M
Run Growth:
-1.2M
Growth Rate:
-11.22%
Total runs:
5.1M
Run Growth:
253.0K
Growth Rate:
4.95%
Total runs:
1.1M
Run Growth:
-333.6K
Growth Rate:
-29.31%
Total runs:
807.4K
Run Growth:
-564.8K
Growth Rate:
-70.07%
Total runs:
276.4K
Run Growth:
-546.4K
Growth Rate:
-200.20%
Total runs:
220.2K
Run Growth:
-1.2M
Growth Rate:
-532.53%
Total runs:
202.8K
Run Growth:
48.8K
Growth Rate:
22.30%
Total runs:
164.5K
Run Growth:
-205
Growth Rate:
-0.12%
Total runs:
50.3K
Run Growth:
-103.2K
Growth Rate:
-200.32%
Total runs:
16.8K
Run Growth:
911
Growth Rate:
5.42%
Total runs:
10.6K
Run Growth:
-13.9K
Growth Rate:
-129.71%
Total runs:
8.7K
Run Growth:
319
Growth Rate:
3.66%
Total runs:
8.7K
Run Growth:
437
Growth Rate:
5.04%
Total runs:
3.8K
Run Growth:
-2.9K
Growth Rate:
-75.21%
Total runs:
3.6K
Run Growth:
2.7K
Growth Rate:
74.30%
Total runs:
3.0K
Run Growth:
590
Growth Rate:
18.64%
Total runs:
2.5K
Run Growth:
1.6K
Growth Rate:
65.76%
Total runs:
863
Run Growth:
-171
Growth Rate:
-19.75%
Total runs:
751
Run Growth:
-789
Growth Rate:
-111.91%
Total runs:
664
Run Growth:
0
Growth Rate:
0.00%
Total runs:
644
Run Growth:
0
Growth Rate:
0.00%
Total runs:
639
Run Growth:
635
Growth Rate:
100.00%
Total runs:
561
Run Growth:
-683
Growth Rate:
-121.75%
Total runs:
339
Run Growth:
-287
Growth Rate:
-85.67%
Total runs:
273
Run Growth:
-3.9K
Growth Rate:
-1419.41%
Total runs:
199
Run Growth:
63
Growth Rate:
31.34%
Total runs:
193
Run Growth:
83
Growth Rate:
42.56%
Total runs:
179
Run Growth:
32
Growth Rate:
18.08%
Total runs:
159
Run Growth:
46
Growth Rate:
28.57%
Total runs:
158
Run Growth:
18
Growth Rate:
11.39%
Total runs:
158
Run Growth:
52
Growth Rate:
33.12%
Total runs:
157
Run Growth:
27
Growth Rate:
17.65%