Oracle today announced the general availability of HeatWave GenAI, which includes industry’s first in-database large language models (LLMs), an automated in-database vector store, scale-out vector processing, and the ability to have contextual conversations in natural language informed by unstructured content. These new capabilities enable customers to bring the power of generative AI to their enterprise data—without requiring AI expertise or having to move data to a separate vector database. HeatWave GenAI is available immediately in all OCI regions, OCI Dedicated Region, and across clouds and at no extra cost to HeatWave customers.

Leading industry analysts had the following to say about HeatWave GenAI:

“Enterprises are increasingly under pressure to find ways to leverage AI as a competitive advantage, but the question is, how to do it? Databases should be seen as key components, even focal points, of any enterprise AI initiative, including that involving unstructured data. Oracle has responded by building an integrated GenAI/database capability. With HeatWave GenAI’s in-database LLMs and in-database vector store, Oracle offers a comprehensive all-in-one solution. This not only avoids the management and synchronizing issues prevalent with specialized vector databases, but also enables exciting new opportunities for developers and users alike. Customers can gain from combining HeatWave AutoML and HeatWave GenAI to improve accuracy and performance while reducing costs by narrowing the context and LLM compute time. HeatWave GenAI is designed to enable developers and users to create a vector store and use a vector store with LLMs using a simple, straightforward process with no AI expertise required. Altogether, HeatWave GenAI delivers a powerful series of AI innovations that organizations with large unstructured data stores should take a closer look at.”
—Carl Olofson
Research vice president, Data Management Software, IDC

“In a world where AI complexity reigns supreme, it’s refreshing to see an enterprise-grade AI solution that helps organizations accelerate real-world GenAI deployments as opposed to spending cycles trying to wade through the sea of options and wasting precious time trying to integrate everything—and hoping that it will eventually work. With in-database LLMs that are ready to go and a fully automated vector store that’s ready for vector processing on day one, HeatWave GenAI takes AI simplicity—and price performance—to a level that its competitors such as Snowflake, Google BigQuery and Databricks can’t remotely begin to approach.”
—Steve McDowell
Chief Analyst, NAND Research

“There’s vector processing and there’s vector processing done right. Oracle has unleashed HeatWave GenAI onto the world and has delivered relentless vector processing performance that is 30X faster than Snowflake, 18X faster than Google BigQuery and 15X faster than Databricks—at up to 6X lower cost. Clearly HeatWave GenAI didn’t just turn the heat up on the competition—it melted them down. For any organization serious about high performance generative AI workloads, spending company resources on any of these three other vector database offerings is the equivalent of burning money and trying to justify it as a good idea.”
—Ron Westfall
Research Director, The Futurum Group

“HeatWave GenAI’s automated in-database vector store and in-database LLMs greatly simplify how developers build AI applications and users interact with data. In contrast to many other vendors that require multiple steps to create vector stores and to use vector stores with LLMs, HeatWave GenAI requires only one SQL statement for each task and it’s all automated from there. It’s equally simple for users. They can use HeatWave Chat for contextual conversations with their data in natural language. Beyond simplification, these in-database capabilities mean that customers gain performance benefits without GPUs and don’t need to manage a standalone vector database, to move data and keep it synchronized, or to incur associated costs.”
—Richard Winter
CEO, WinterCorp

“HeatWave is taking a big step in making generative AI and Retrieval-Augmented Generation (RAG) more accessible by pushing all the complexity of creating vector embeddings under the hood. Developers simply point to the source files sitting in cloud object storage, and HeatWave then handles the heavy lift. Customers can use Oracle’s embedding model or bring their own. And then they can use in-database AutoML to ground the language model and reduce hallucinations all without having to call separate databases or send the data elsewhere. These are major steps toward democratizing AI.”
—Tony Baer
Founder and CEO, dbInsight

“HeatWave’s engineering innovation continues to deliver on the vision of a universal cloud database. The latest is generative AI done ‘HeatWave style’—which includes the integration of an automated, in-database vector store and in-database LLMs directly into the HeatWave core. This enables developers to create new classes of applications as they combine HeatWave elements. For example, they can combine HeatWave AutoML and HeatWave GenAI in a fraud detection application that not only detects suspicious transactions—but also provides an understandable explanation. This all runs in the database, so there’s no need to move data to external vector databases, keeping the data more secure. It also makes HeatWave GenAI highly performant at a fraction of the cost as demonstrated in competitive benchmarks.”
—Holger Mueller
VP and Principal Analyst, Constellation Research

“With the launch of HeatWave GenAI, Oracle has applied an economic lens to the world of vector processing and, through third party benchmarks, demonstrated that HeatWave outperforms the competition for the workloads tested. This is relevant for executives responsible for paying monthly GenAI cloud bills as the more efficiently you can perform vector processing workloads, the less you spend. We’re intrigued by what we see as industry-first, in-database LLMs and an automated vector embedding capability. Add in the HeatWave Lakehouse solution to help organizations looking to get started in AI, as well as those ready for large-scale deployments, and you can see the Heatwave train continues to roll.”
—Dave Vellante 
Co-CEO and Chief Analyst of theCUBE Research

“Database vendors have been jumping on the GenAI, LLM, and vector search trend by adding vector capabilities to their existing databases. Customers are finding most of these implementations are too slow, too costly, or both. Oracle is changing the game. Based on a new independent benchmark, Snowflake, Databricks and Google BigQuery are several times slower at all aspects of vector processing – embedding, loading, and querying – than HeatWave GenAI. In fact, the benchmark results show that HeatWave GenAI is anywhere from 15X-30X faster, and as much as 6X cheaper. That’s like getting the performance of a Rimac Nevera for the price of a Kia EV6. That’s an incredible HeatWave GenAI value proposition. For organizations looking to run highly accurate and blazingly fast similarity search queries on unstructured data in object storage at an affordable price, HeatWave has the answer today.”
—Marc Staimer
Senior Analyst, theCUBEresearch

“With the latest release of HeatWave GenAI, Oracle has again pushed the boundaries of database technology, demonstrating 30x better performance than the competition at a substantially lower cost. More important, however, is that customers now have a fully integrated GenAI platform that includes everything from a vector store and in-database LLMs to machine learning and natural language capabilities under the same hood. This approach eliminates unnecessary data movement and fragmentation, reduces the attack surface area, and helps to dramatically improve regulatory compliance posture. HeatWave GenAI exemplifies elegant and synergistic engineering design, demonstrating that such a combination of performance, efficiency, and security is impossible to achieve by just randomly connecting individual cloud services.”
—Alexei Balaganski
Lead Analyst and CTO, KuppingerCole Analysts

“Generative AI holds a world of promise to organizations of all types and sizes. With that big promise comes big challenges as IT organizations are challenged in terms of budget, technology, and AI skills. Because of this, many organizations find themselves having to pause generative AI projects, putting them at risk of falling behind the competition. With HeatWave GenAI, IT organizations can easily move from concept to production through in-database large language models (LLMs) and vector store. This enables IT and app developers to quickly embed GenAI into its most precious resource – its data. These capabilities, married with HeatWave Chat also enable users and customers to perform queries and searches through natural conversations.”
—Matt Kimball
Vice President and Principal Analyst, Moor Insights & Strategy.

“What do you get when you combine in-database LLMs and an automated vector store with a scale-out data processing engine optimized for data in object store? You get Oracle’s newly released HeatWave GenAI. In typical Oracle fashion, this release seeks to outperform the company’s primary competitors such as Snowflake. In terms of speeding up vector processing, Oracle is claiming a rather significant factor of 30X over its competitors “With or without such performance assertions, the bottom line is that HeatWave GenAI packs some serious performance “heat” by doing similarity search processing in-memory. This goes beyond what we’ve seen from early vector databases and what customers get from many vector “plug-ins” for traditional databases. Importantly, HeatWave GenAI brings this performance mentality to bear on development as well as execution by streamlining database setup and even automating complex tasks like model selection and content parsing/chunking. These moves will speed up the creation of impactful GenAI applications including applications like content summarization, retrieval augmented generation, and natural language conversations with data.”
—Bradley Shimmin
Chief Analyst, AI & Data Analytics, Omdia

Lear more about HeatWave GenAI and try it out!