H2O.ai
Free Apache-2.0 AutoML engine, plus a quoted enterprise platform for private AI
Overview
H2O.ai offers H2O-3, a free Apache-2.0 distributed machine-learning library with AutoML, alongside a commercial platform for regulated industries: H2O-3 Secure, Driverless AI automated modelling and Enterprise h2oGPTe for private generative and agentic AI. Commercial products run in the cloud, on premises or air-gapped and are sold on an annual licence priced by compute and memory.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 14 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- All H2O-3 algorithms and H2O AutoML
- Multi-node distributed model training and model scoring
- PyPI and R-CRAN packages
- Community support
- Automated model documentation
- Enterprise SSO, RBAC and audit
- Production MOJO scoring for Hadoop, Spark, Teradata and Oracle
- Premium support with SLAs and commercial CVE patching
- Driverless AI automated feature engineering, tuning and model documentation
- Enterprise h2oGPTe with agentic RAG, document AI and the Super Agent
- Cloud, on-premises or air-gapped deployment
What you can produce with H2O.ai
- Trained models from the free H2O-3 engine, including an AutoML leaderboard
- Automated feature engineering, model selection and tuning in Driverless AI
- Model documentation generated automatically as each model trains
- Explanations of model behaviour with Shapley values, K-LIME and surrogate trees
- Cited answers over company documents from Enterprise h2oGPTe
- Agents that plan and run multi-step tasks through the H2O AI Super Agent
- Private deployment on premises, in a private cloud or air-gapped
H2O-3 remains one of the most capable free machine-learning engines available: distributed training, a sensible AutoML baseline and Python, R and JVM interfaces under Apache 2.0, with no usage cap. Production extras such as automatic model documentation, SSO and CVE patching now sit in the paid H2O-3 Secure edition.
The commercial platform, Driverless AI plus h2oGPTe and its agents, is built for banks, telcos and government bodies that must keep data on their own infrastructure, with FedRAMP High and IMDA accreditation to show for it. It is sold only by quote on an annual compute-based licence, so small teams should stay on the open-source engine.
What we found
H2O.ai sells two things under one name, and they suit very different buyers. The first is H2O-3, an open-source distributed machine-learning engine under the Apache 2.0 licence, with gradient boosting, GLMs, deep learning, stacked ensembles and an AutoML routine that H2O says has shipped since 2017. It installs with pip or from CRAN, runs from Python, R, Java or Scala, and was still getting releases in August 2026. H2O says more than 20,000 organisations use the open-source engine; that is the company's figure.
The second is a commercial platform for regulated industries. Driverless AI automates feature engineering, model selection, tuning and model documentation; Enterprise h2oGPTe handles retrieval with citations, document AI and the H2O AI Super Agent across hosted or private models. In 2026 H2O added tabH2O, a foundation model for tabular data, earned FedRAMP High certification for its government cloud, renewed its IMDA accreditation in Singapore and expanded a forward-deployed engineering lab there. The older open-source h2oGPT repository is now archived.
A middle option is easy to miss. H2O-3 Secure is the same code with automatic model documentation, SSO and role-based access, CVE patching, production scoring packages and support under a service-level agreement. Those features are now held back from the free edition.
Pricing is the obstacle for small teams. No price is published anywhere; H2O says it charges one annual licence based on compute and memory rather than per token, and every route leads to a demo request. One AWS Marketplace reviewer in November 2020 liked the AutoML for quick prototypes but wrote that Driverless AI "may not be affordable to the small fish in the pond". H2O raised a $100 million Series E in November 2021 at a $1.6 billion valuation, led by the Commonwealth Bank of Australia, now also a customer. Start with H2O-3, and talk to sales only once governance or private generative AI becomes the requirement.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
About this listing
This entry was compiled from publicly available data including H2O.ai's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with H2O.ai unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to H2O.ai directly →
Spotted something out of date? Suggest an update →
Alternatives to H2O.ai
More in Code & Dev Tools
