cornserve.mp4
Project News
- [2026/04/15] Our paper ("Cornserve: A Distributed Serving System for Any-to-Any Multimodal Models") has been accepted to the ACM CAIS 2026 Demo Track! Paper
- [2025/12/16] Our planner paper preprint ("Cornfigurator: Automated Planning for Any-to-Any Multimodal Model Serving") is now available on arXiv.
- [2025/11/14] Cornserve project announced with v0.1.0 release!
Cornserve is a distributed inference platform for complex, Any-to-Any multimodal AI. Split complex models into smaller separately scalable components (model fission) and share common components across multiple applications (sharing), all on your own infrastructure.
See also Cornfigurator, an automated deployment planner for Cornserve.
You can quickly try out Cornserve on top of Minikube. Check out our getting started guide!
Cornserve can be deployed on Kubernetes with a single command. More on our docs.
These research papers describe Cornserve's system architecture and planner.
- Cornserve (CAIS 26): Paper
- Cornfigurator: Paper | Repository
If you find Cornserve relevant to your research, please consider citing:
@inproceedings{cornserve-cais26,
title = {Cornserve: A Distributed Serving System for Any-to-Any Multimodal Models},
author = {Chung, Jae-Won and Ma, Jeff J. and Ahn, Jisang and Liang, Yizhuo and Jajoo, Akshay and Lee, Myungjin and Chowdhury, Mosharaf},
booktitle = {ACM CAIS},
year = {2026}
}
@article{cornfigurator-arxiv25,
title = {Cornfigurator: Automated Planning for Any-to-Any Multimodal Model Serving},
author = {Ma, Jeff J. and Chung, Jae-Won and Ahn, Jisang and Liang, Yizhuo and Lu, Runyu and Jajoo, Akshay and Lee, Myungjin and Chowdhury, Mosharaf},
journal = {arXiv preprint arXiv:2512.14098},
year = {2025}
}Cornserve is an open-source project, and we welcome contributions! Please check out our contributor guide for more information on how to get started.