The deployment layer around a model
BentoML names both an open-source model-serving framework and the team that built commercial inference tooling around it. Its Python framework packages models and application logic into services, with tools for batching, streaming, distributed execution, observability and deployment. BentoCloud adds managed deployment and scaling. [2]
These tools address the work between a trained model and a service that reliably handles requests. A pipeline may combine several models or preprocessing steps; packaging and deployment make that pipeline callable and repeatable. Supporting a framework or model format does not make every compatible model developer a corporate partner. [2]
A product acquisition and a continuing project
On February 10, 2026, BentoML's team joined Modular. BentoML called the transaction a strategic product acquisition, while Modular described adding managed deployment infrastructure to its MAX and Mojo stack. The announcements preserved the BentoML open-source project under Apache 2.0. [1][3]
Transaction counsel published same-day confirmation. The record establishes a completed product and team combination, while the companies' announcements do not specify a purchase of all shares in the BentoML legal entity. This profile therefore describes the integration without asserting a corporate parent. [1][4]
Sources
- BentoML Is Joining Modular
BentoML · Feb 10, 2026
- BentoML Documentation
BentoML · retrieved Oct 23, 2026
- BentoML Joins Modular
Modular · Feb 10, 2026
- SLS Client BentoML is Acquired by Modular
Silicon Legal Strategy · Feb 10, 2026