Oracle Cloud Infrastructure (OCI) AI updates have recently focused on a common theme: making it easier for customers to move from experimenting with AI to putting it to work. From the launch of OCI Enterprise AI and the availability of cutting-edge models like Grok 4.3 and NVIDIA Nemotron 3 Nano Omni, to real-world deployments like SoftBank’s sovereign AI platform and prebuilt AI Accelerator Pack solutions that speed up time-to-value, these recent updates are all about reducing complexity while increasing flexibility and control. Together, they show how OCI is helping organizations build, scale, and operationalize AI faster—without compromising on performance, governance, or where their data lives.
New updates and features:
xAI Grok 4.3 now available on OCI
Just one day after its public release, Grok 4.3 is now available on OCI Enterprise AI, bringing xAI’s latest reasoning model to OCI with strong performance across advanced logic, math, coding, and multi-step analysis (98% on τ²-Bench Telecom, 81% on IFBench), plus a one million-token context window.
It also ranks in the global top 10 on the Artificial Analysis Intelligence Index, sits on the Pareto frontier for intelligence per dollar, and delivers strong agentic performance at a significantly lower cost than comparable frontier models.
NVIDIA Nemotron 3 Nano Omni on OCI
NVIDIA Nemotron 3 Nano Omni is now available on OCI Enterprise AI, bringing a fully open-source, multimodal model that can reason across video, audio, images, and text in a single system, making it easier to build more advanced AI applications without stitching together multiple models. On top of that, the open-source model combined with OCI’s dedicated AI clusters gives teams control over how and where they deploy, customize, and run AI.
Learn more about Nemotron 3 Nano Omni
SoftBank’s Sovereign AI Platform, Powered by OCI
SoftBank is using OCI’s AI services to power its new sovereign cloud platform in Japan, combining its own generative AI models with Oracle Cloud Infrastructure and OCI Enterprise AI to deliver advanced AI capabilities while keeping data fully controlled within local data centers. It’s a great example of how enterprises can build and scale AI on OCI without giving up control over security, governance, or data residency—something that’s especially important in regulated industries. By pairing OCI’s full stack of AI services with its OCI Alloy, SoftBank is able to offer customers powerful AI tools alongside strict data sovereignty, showing how OCI can support real-world, large-scale AI deployments in highly regulated environments.
New resources for AI at OCI:
Webinar replay – Turning intelligence into action with OCI Enterprise AI
Last month, Oracle launched OCI Enterprise AI, a new platform that helps customers build, deploy, and scale AI solutions with less complexity and more confidence. OCI Enterprise AI experts hosted a webinar introducing the new offering and provided insight into the product capabilities, benefits, and real-world use cases so customers can jump right in and start seeing value right away.
6 Ways to Apply Prebuilt AI Solutions for Faster Time-to-Value
Discover six practical ways to apply OCI AI Accelerator Packs to common enterprise challenges across support, sales, compliance, and operations. These prebuilt solutions combine infrastructure, models, and workflows to help teams move from idea to production faster.
Many AI initiatives stall not because of their intended use case, but because teams must assemble data pipelines, models, and infrastructure before they can begin to see value. OCI AI Accelerator Packs provide a complete, deployed AI system tailored to specific business scenarios, reducing integration effort and accelerating time to impact. This approach helps organizations focus on outcomes instead of building and maintaining the underlying stack.
Three ways Oracle AI transforms document processing
Let AI turn document overload into a competitive advantage. Oracle Cloud Infrastructure (OCI) Document Understanding uses AI-powered generative extraction to eliminate manual processing—extracting data across formats, accelerating workflows, and driving faster, smarter decisions.Generative extraction turns complex, unstructured documents into AI-ready data without relying on rigid templates, manual labeling, or ongoing retraining. By understanding document context, structure, and relationships, it can extract key information more accurately across varied formats, including tables, charts, and freeform content. This reduces errors, rework, and maintenance costs while enabling faster automation and downstream AI workflows.