The new OCI Generative AI service is fully managed, cloud-based and provides access to Llama 2 and other powerful LLMs.

Oracle launches its highly anticipated Oracle Cloud Infrastructure Generative AI service. This allows companies to tackle large language models (LLM) like Metas Llama 2. In total, the service supports more than 100 languages. The portal also includes improved cluster management for GPUs and other fine-tuning options.

According to Oracle, the new service provides a foundation for these models, which can then be used in services such as text generation, summarization and semantic similarity.

When announcing the service in the middle of last year, Oracle described the platform as an alternative to Microsoft’s Azure OpenAI service. This has long allowed you to introduce LLMs based on human input through conversations.

You can use existing AI models from Llama 2 or other languages, but everything becomes much more powerful when you add your own data. You can add this precisely using the OCI Generative AI Agents service. This feature is currently in beta and provides access to a RAG agent to leverage the search capabilities built into OCI Open Search. This means you always receive the most up-to-date information.

In the long term, Oracle also plans to add other data search and aggregation tools in addition to RAG, including Oracle Database 23c, AI Vector Search and MySQL HeatWave with Vector Store.

The OCI Generative AI service is available on Oracle Cloud in various regions. If you want to dive deeper into the different options, you can visit Oracle’s blog.