RepoStepFunStepFunpublished Jul 25, 2025seen 5d

stepfun-ai/Step3

Open original ↗

Captured source

source ↗
published Jul 25, 2025seen 5dcaptured 9hhttp 200method plain

stepfun-ai/Step3

License: Apache-2.0

Stars: 453

Forks: 12

Open issues: 5

Created: 2025-07-25T03:11:20Z

Pushed: 2025-08-10T15:26:18Z

Default branch: main

Fork: no

Archived: no

README:

Introduction

Step3 is our cutting-edge multimodal reasoning model—built on a Mixture-of-Experts architecture with 321B total parameters and 38B active. It is designed end-to-end to minimize decoding costs while delivering top-tier performance in vision–language reasoning. Through the co-design of Multi-Matrix Factorization Attention (MFA) and Attention-FFN Disaggregation (AFD), Step3 maintains exceptional efficiency across both flagship and low-end accelerators.

Step3 model card:

| Config | Value | |------------------------|---------| | Number of Layers (Dense layer included)|61| |Number of Dense Layers| 5| | Hidden Dimension | 7168 | | Attention Mechanism | MFA | | Low-rank Query Dimension | 2048 | | Number of Query Heads | 64 | | Head Dimension | 256 | |Number of Experts |48| |Selected Experts per Token|3| |Number of Shared Experts| 1| | Max Context Length | 65536 | | Tokenizer | Deepseek V3 | | Total Parameters (LLM) | 316B | | Activated Params per Token | 38B | | Total Parameters (VLM) | 321B |

Evaluation Results

![](figures/step3_bmk.jpeg)

Deployment

> You can access Step3's API on https://platform.stepfun.com/ , we provide OpenAI/Anthropic-compatible API for you. >

Our model checkpoints are stored in bf16 and block-fp8 format, you can find it on Huggingface.

Currently, it is recommended to run Step3 on the following inference engines:

  • vLLM
  • SGLang

Deployment and Request examples for vLLM and SGLang can be found in the [Model Deployment Guide](docs/deploy_guidance.md).

Contact Us

If you have any questions, please reach out at [contact@stepfun.com](mailto:contact@stepfun.com) .

License

Both the code repository and the model weights are released under the [Apache License (Version 2.0)](./LICENSE).

Citation

@misc{step3system,
title={Step-3 is Large yet Affordable: Model-system Co-design for Cost-effective Decoding},
author={StepFun Team},
year={2025},
eprint={2507.19427},
archivePrefix={arXiv},
primaryClass={cs.LG},
url={https://arxiv.org/abs/2507.19427},
}

@misc{step3blog,
title={Step3: Cost-Effective Multimodal Intelligence},
author={StepFun Team},
url={https://stepfun.ai/research/step3},
}

Notability

notability 6.0/10

New model repo with moderate traction.