Model Serving Made Easy
From trained ML models to production-grade prediction services with just a few lines of code
A simple yet flexible workflow empowering
Data Science teams to continuously ship prediction services
Unified format for deployment
High performance model serving
100x the throughput of your regular flask based model server, thanks to our advanced micro-batching mechanism. Read about the benchmarks here.
DevOps best practices baked in
Organizations using and contributing to BentoML
Built to work with DevOps & Infrastructure tools
KNative
CloudRun
BentoML for teams (beta)
Serving Admin
Fully managed
Copyright 2020 BentoML
Keep all your team's models, deployments, and changes highly visible and control access via SSO, RBAC, Client Authentication, and auditing logs.
DevOps-free BentoML workflow, from prediction service registry, deployment automation, to endpoint monitoring, all configured automatically for your team. A solid foundation for running serious ML workloads in production.
Deliver high quality prediction services that speaks the DevOps language and integrates perfectly with common infrastructure tools.
Unified model packaging format enabling both online and offline serving on any platform.
BentoML supports all major ML frameworks
Built with BentoML
Learn more about BentoML
Featured
Sentiment analysis with BERT
Image Classification
The service uses the BERT model trained with the TensorFlow framework to predict movie reviews' sentiment.
Titanic Survival Prediction
This service uses ResNet50 from ONNX model zoo to identify objects in a given image.
This prediction service use model trained with XGBoost framework to predict surivival rate of giving passenger on the Titanic cruise ship.
Core ML
