Auto-PyTorch vs BigML
Auto-PyTorch
automl.github.ioBigML
bigml.comSide by side
| Figure | Auto-PyTorch | BigML |
|---|---|---|
| Price from | Free | $833.33/mo |
| Free plan | Yes | Yes |
| Free trial | — | Yes |
| Open source | — | — |
| Web | — | Yes |
| macOS | — | Yes |
| Linux | Yes | — |
| API | — | Yes |
| Self-hosted | Yes | Yes |
| Plans published | 1 | 7 |
| Spec score · AutoML Software | 8.4 (#5) | 9.1 (#1) |
| Deployment | The project provides a Docker image and can be installed from PyPI or manually in a Python environment. | Private deployments can run on a preferred cloud provider, ISP, or on-premises behind a corporate firewall, as managed or self-managed deployments. |
| Integrations | The documentation describes using Dask.distributed for parallel Bayesian optimization and sklearn column transformers for data preprocessing. | Official tools include Google Sheets, Zapier, Node-RED, Alexa, PredictServer and a native BigMLX macOS application. |
| What it does | Auto-PyTorch is an automated machine-learning toolkit based on PyTorch that helps users automate algorithm selection and hyperparameter tuning. | BigML provides cloud and deployable machine-learning workflows for classification, regression, forecasting, clustering, anomaly detection, association discovery and topic modeling. |
Plans
BigML
- FREEFree
- Basic BigML training$800, paid once
- BigML Lite$1,000 a month
- Extra Support$3,000 a month
- BigML Lite$10,000 a year
- BigML Lite (annual)$10,000 a year
- Bronze Enterprise$45,000 a year
In their own words
Auto-PyTorch
Auto-PyTorch is a free, self-hosted machine-learning toolkit based on PyTorch. It automates algorithm selection and hyperparameter tuning, including neural-network architecture and training settings, within a user-defined search budget. Documented search methods include Bayesian optimisation, meta-learning, SMAC and Hyperband. Supported work includes tabular classification, tabular regression and time-series forecasting. Tabular preprocessing can perform imputation, categorical encoding, scaling and feature processing, while related hyperparameters are tuned. The toolkit can build ensembles from validation-set predictions, with controls for ensemble size and candidate limits. Users can set estimator memory limits and total wall time. Parallel Bayesian optimisation uses Dask.distributed; workers need shared filesystem access to training data and models. Installation documentation specifies Linux, Python 3.7 or later, a C++11-capable compiler and SWIG 3.0. SWIG 4.0 or later is unsupported. Forecasting needs extra dependencies. A Docker image and PyPI installation are documented, under the 3-clause BSD license.
BigML
BigML is a machine-learning platform for classification, regression, forecasting, clustering, anomaly detection, association discovery, and topic modeling. It provides interactive model visualizations, prediction explanations, and field-importance views. Models can be exported in JSON PML or PMML formats for use in programming languages and web, mobile, or IoT applications. OptiML helps select and tune models; WhizzML automates workflows, and Scriptify turns workflows into reusable scripts. Users can work through the REST API and language bindings including Python, Node.js, Ruby, Java, and Swift. Listed integrations include Node-RED, Google Sheets, Zapier, Alexa, Docker, and macOS. BigML runs on web, macOS, API, and self-hosted platforms, with private deployments available on a cloud provider, ISP, or on-premises. The free plan is $0/month and includes up to 60 tasks, a 16 MB per-task dataset limit, and two parallel tasks. Paid plans start at $1,000/month; a three-day trial is available.