Oracle Autonomous AI Lakehouse

Overview

The service queries Apache Iceberg tables in place across clouds, so teams can work with lake data without relocating it. It combines this lake access with SQL analytics, transactions, streaming ingestion, and support for structured, semi-structured, and unstructured data. Separate storage and compute and hybrid deployment give teams flexibility in how they arrange workloads.

Oracle says the service runs on OCI, AWS, Azure, Google Cloud, and Exadata Cloud@Customer. That range suits teams with data and infrastructure across providers, while the Oracle database foundation makes it a stronger fit for organizations already invested in Oracle than for buyers seeking only a focused lake query layer.

Key features

Lake access and data movement

Catalog integrations include OCI Data Catalog, AWS Glue, and Apache Iceberg catalogs in Databricks and Snowflake. Data Studio adds drag-and-drop workflows for integrating data from more than 100 application, cloud service, and database sources. Those options can reduce integration work for teams using these ecosystems, though the breadth matters less if their workflows are already served by a narrower tool.

Data Studio also supports bidirectional sharing with services including Power BI and Tableau through Delta Sharing. This is useful when teams need to exchange data with those analytics tools rather than funneling every workflow through a single environment.

AI, analytics, and operations

Select AI can use models from Cohere, Azure OpenAI, OpenAI, OCI Generative AI, Google, Anthropic, Hugging Face, and AWS, among others. Machine learning, graph analytics, spatial features, and AI Vector Search extend the service beyond SQL analysis to semantic search and retrieval-augmented generation. The range is attractive for teams that want these workloads near their database, but buyers focused on a single AI use case may not need such a broad platform.

Oracle says autonomous management handles provisioning, configuration, security, tuning, and scaling. Autonomous AI Database encrypts data at rest and in transit by default and applies security patches and updates automatically; Oracle also states that it meets a broad set of international and industry-specific compliance standards. These capabilities reduce routine operational and security work, but teams should still assess their own workload and compliance requirements.

Local development and document data

An unlimited-time container image supports offline local development with Database Actions, ORDS, APEX, and the Database API for MongoDB. JSON and OSON documents are supported, with a maximum document size of 32 MB. This gives developers a local path for working with the database tools, though it is distinct from the service’s cloud usage and billing options.

Pricing

The pricing model is freemium. The Always Free Autonomous AI Lakehouse plan costs 0.00 USD per free, billed as free usage for an unlimited time, subject to capacity limits. It includes two Autonomous AI Database instances through Oracle Cloud Free Tier. That makes it a low-commitment way to explore the service or support small workloads, but capacity limits mean it is not a dependable substitute for paid capacity where workload needs are fixed.

Oracle Autonomous AI Lakehouse Serverless has custom pricing, billed per ECPU per hour, with storage and backup storage billed per gigabyte per month. The pricing page lists ECPU, storage, backup storage, and a developer instance. This usage-based structure suits teams that prefer a serverless option, but they need to account for both compute and storage consumption.

Oracle Autonomous AI Lakehouse on Exadata Cloud@Customer and Oracle Autonomous AI Lakehouse on Dedicated Infrastructure also have custom pricing, billed per ECPU per hour and per developer instance per hour. The dedicated infrastructure database subscription has a 48-hour minimum term, which matters for teams planning short-lived usage. Bring Your Own License is available for Serverless, Dedicated, and Exadata Cloud@Customer and is billed per ECPU per hour; it is relevant to organizations bringing an existing license, though the ECPU charge remains.

Oracle offers US$300 in cloud credits for up to 30 days. Credits expire when spent or when 30 days elapse, whichever comes first, so the trial is a bounded evaluation rather than ongoing free capacity.

Platforms

The platform supports API, Linux, self-hosted, and web environments. Alongside managed cloud availability, Exadata Cloud@Customer and hybrid deployment give organizations options that include customer infrastructure.

Who it's for

Oracle Autonomous AI Lakehouse is a strong candidate for teams that already use Oracle databases and want to query Iceberg data in place while adding integration, sharing, and AI capabilities. Its cloud coverage, catalog connections, and multiple model providers also suit organizations with varied infrastructure and analytics stacks.

It is less compelling for teams that need only a lake query engine, want predictable costs without monitoring usage, or cannot work within the free plan’s capacity limits. In those cases, the broader database platform may add complexity without enough benefit.

Pros and cons

• Pros: Queries Iceberg tables in place across clouds, avoiding data movement for that access pattern.
• Pros: Integrates with multiple catalogs and more than 100 data sources, and supports bidirectional sharing with Power BI and Tableau.
• Pros: Combines database analytics, vector search, machine learning, graph and spatial features, and a wide selection of AI model providers.
• Pros: Free use has no time limit, and the local container image supports offline development.
• Cons: Free usage is subject to capacity limits, so it may not support steady production needs.
• Cons: Paid choices use ECPU-based billing, with additional storage and backup charges on Serverless; the Dedicated Infrastructure subscription also has a 48-hour minimum.
• Cons: The breadth of Oracle database and AI features may be unnecessary for teams seeking only a focused lake query service.

Alternatives

Starburst Data Platform is worth considering for teams that want a free tier capped at three clusters and standard cluster execution for ad hoc queries.

Starburst Galaxy also offers a free tier with up to three clusters and standard execution for ad hoc queries; it is an option for teams seeking that bounded cluster-based starting point.

Bauplan may fit teams that prefer a shared sandbox with public datasets or their own uploads, a full CLI, SDK, and API, and community support, with the caveat that everything in the sandbox is public.

IOMETE is a relevant alternative for teams seeking self-hosted, on-premises deployment: its free plan is capped at 100 vCPUs and includes core features and community support.

Cloudera Data Lake Service is another option for buyers evaluating a paid data lake service.

Amazon SageMaker Autopilot suits readers focused on a pay-as-you-go machine-learning service; its on-demand pricing has no minimum fees or upfront commitments.

Databricks Notebooks offers a free edition with one serverless workspace and limited compute size and usage, which may suit notebook-centered evaluation.

Dremio offers a forever-free tier with three projects, though customer cloud infrastructure costs may still apply.

For broader comparisons, browse Data Lakehouse Platforms, Data Warehouse Software, Document Databases, OLAP Databases, and OLAP Software.

Verdict

Choose Oracle Autonomous AI Lakehouse if your team wants Oracle database capabilities, in-place Iceberg queries, and integrated data engineering and AI across a mixed-cloud stack. Its breadth is the main reason to buy; capacity-constrained free use and consumption-based paid billing are the main reasons to look elsewhere.