A faster way to ship your models to production

BentoML combines the best developer experience with a focus on operating ML in production. 

Our platform enables Data Science teams to do their best work.

Sign UpCommunityDocs

Get started with BentoML today!

Open Source

Scale whenever needed

Collaboration is better

Copyright © 2022 BentoML

GalleryCommunityDocumentationTutorialContactOpen Source
Try BentoML Now

$ pip install bentoml

$ helm repo add yatai https://bentoml.github.io/yatai-chart
Blog

BentoML's open source model serving framework is blazing fast and easy to get started. Check out the tutorial and documentation for next steps.

An open platform for ML in production

BentoML is compatible across machine learning frameworks and standardizes ML model packaging and management for your team.

Python-first, scales with powerful optimizations

Maximize resource utilization across multi-stage serving pipelines.

Parallel Inference

Adaptive Batching

Dynamically group predictions requests in real-time into batches for model inference.

Accelerated Runtime

Run your model serving workloads seamless with accelerated hardware.

Yatai: Model Deployment at scale on Kubernetes

Yatai provides an open and flexible MLOps platform that streamlines the collaboration between Data Science and Engineering teams.

Yatai is a cloud native platform built to leverage BentoML and Kubernetes, offering ML teams a future-proof foundation for scaling their team or ML workload.

Ready to get started?

Sign up for BentoML Cloud today and learn about how we can help accelerate your machine learning projects to production.

SIGN UPSCHEDULE A DEMOJoin the 🍱 community !👉 Try out Yatai open source

BentoML integrates with your existing data stack and enable your ML team to continuously deploy and improve their models in production.