Who it is built for
- ML teams training large AI models
- Developers using CLI and Python SDK workflows
- Organizations running pretraining or fine-tuning jobs
Train and deploy large AI models with Databricks
Databricks Mosaic AI Training is the current Databricks service and tooling for training and deploying large AI models. Formerly MosaicML, it provides a command-line interface and Python SDK for training runs, pretraining, fine-tuning, cloud data access, and developer-tool integrations.
The decision
Start with the job, the team, and the constraints. Product fit becomes much clearer when those three line up.
Inside the product
The core product capabilities, grouped around the work they enable.
The platform provides a command-line interface and Python SDK for configuring and submitting training runs.
Its documentation covers workflows for pretraining and fine-tuning supported models.
It supports connections with Git, Docker, remote cloud storage, and experiment-tracking tools.
After onboarding, users authenticate in the Mosaic AI console, configure the CLI with an API key, and submit training runs. They can set up remote-cloud credentials and use YAML run configurations or the Python SDK to train, pretrain, or fine-tune supported models.
Plans and official links
See the entry price, free access options, company details, and direct vendor destinations in one place.
Choosing for a real workflow?
We map the workflow, connect existing systems, choose what to buy, and build what is missing.