We're upgrading the model backbone behind Planning Analytics Workspace's AI capabilities — bringing better performance, improved outputs, broader regional availability, expanded language support and a stronger foundation for the road ahead.

At IBM, we're committed to continuously improving the AI experience inside IBM Planning Analytics. Today, we're excited to share a significant under-the-hood upgrade that directly benefits every customer using the IBM-managed AI capabilities in Planning Analytics as a Service — available on both AWS and Azure – these changes are also coming soon for the other deployment options.

From IBM Granite 4 to GPT OSS 120B on Amazon Bedrock

In Planning Analytics Workspace 3.1.10, we have transitioned the AI model powering our managed AI capabilities from IBM Granite 4 models on IBM watsonx.ai to GPT OSS 120B models on Amazon Bedrock. This is a substantial leap in model capability and represents our commitment to delivering the best possible AI outputs to our customers — regardless of which underlying platform provides them.

What does this mean for you? Richer, more accurate, and more contextually relevant AI responses across every AI-powered feature in Planning Analytics Workspace — with no configuration changes required from you.

This change applies specifically to the IBM-managed experience of the Planning Analytics Agent — this is where you bring the data and we bring the AI. The following in-app capabilities are all powered by this upgraded model.

You don't need to be a data scientist or AI engineer to benefit from this update. These capabilities are embedded directly into your Planning Analytics Workspace experience and are ready to use today with IBM Planning Analytics as a Service.

Expanded Regional Availability

One of the most impactful outcomes of this move to Amazon Bedrock is significantly expanded regional support. IBM Planning Analytics Agent deployments on Planning Analytics as a Service are now available in three major geographic regions:

This means teams across the globe — from finance offices in Sydney to FP&A teams in Frankfurt and New York — can now take advantage of the full IBM Planning Analytics Agent experience with the performance and data residency expectations their organizations require.

Expanded Language Support

In addition to expanded regional availability, this change in foundation of the Planning Analytics Agent expands language Support for both the AI Agent chat as well as the AI features of the managed Planning Analytics Agent like Chart insights, Explain cell and more.

Who Gets This Update?

For our on-premises customers, this update will be coming in a future release of Planning Analytics, and we'll be sharing specific details and timelines soon. Stay tuned to our release notes and announcements.

Part of a Larger AI Strategy

This model upgrade is not a one-off improvement — it is a deliberate step in a broader strategy to make IBM Planning Analytics Agent the most capable and flexible AI layer for enterprise planning.

Our vision is to give every customer a clear on-ramp to AI-powered planning, whether they are just getting started or building sophisticated custom workflows:

Managed Experience

Start immediately, no AI expertise needed

The IBM-managed experience gives you powerful AI capabilities out of the box. Chart Insights, Variance Analysis, Explain Cell and the Planning Analytics Agent are all ready to use the moment you log in. We manage the models, infrastructure, and updates — you focus on the business.

Custom Experience

Build your own agents with MCP tools

For teams who want to go further, the Planning Analytics Agent add-on also includes our Model Context Protocol (MCP) tools. These tools allow you to build custom AI agents that are deeply integrated with your Planning Analytics data, workflows, and business logic.

The path is clear: start with the managed experience on day one and expand into fully custom AI agents as your needs evolve — all within the IBM Planning Analytics ecosystem you already trust.

What Should You Do Next?

If you are on Planning Analytics as a Service or Planning Analytics on Cloud, no action is required as we roll out PAW 3.1.10 and PAW 2.1.23. Simply open Planning Analytics Workspace and start exploring the improved AI outputs through Chart Insights, Variance Analysis, Explain Cell etc.

If you have deployed Planning Analytics on-premises, watch this space. We are committed to bringing this improvement to you and will communicate timelines as soon as they are confirmed.

If you haven't yet explored the Planning Analytics Agent add-on or the MCP tooling, now is a great time to start. Reach out to your IBM account team or visit IBM Planning Analytics documentation to learn more.

The IBM Planning Analytics team is committed to making enterprise planning smarter, faster, and more accessible. Updates like this are a direct reflection of our investment in your success. We look forward to sharing more milestones in the months ahead.