Pegasystems has shipped Pega Infinity 26, the general availability release of its enterprise AI suite, with a pricing architecture that charges per completed case rather than per token consumed. The company frames this as eliminating what it calls the "AI token tax" — the unpredictable consumption spend that accumulates when AI agents call language models repeatedly through operational workflows.
The Predictable AI architecture underpinning Infinity 26 is designed to front-load AI work at design time, so that runtime execution draws on pre-computed logic and deterministic rules rather than making fresh model calls for every interaction. The result, according to Pega, is token-efficient runtime performance that keeps actual model costs low enough to support a flat-fee commercial model.
"The next phase of AI adoption is about scaling results with confidence," said Kerim Akgonul, chief product officer at Pega. "Pega Infinity 26 removes two of the biggest barriers to success — unpredictable outcomes and unpredictable costs — by building governance into AI-driven workflows and aligning pricing with completed work."
The release adds three capability areas beyond the pricing change. Pega Infinity Studio is a new AI-native development environment that combines clients' coding agents with built-in best practices from Pega Blueprint AI, targeting faster delivery of mission-critical applications. The platform now supports Model Context Protocol at both design and runtime, meaning teams can invoke Pega agents and workflows from external AI tools including Claude Code, GitHub Copilot, OpenAI Codex, and Microsoft Copilot for Office 365. Customer Engagement Studio is a new agent-powered workspace that automates marketing campaign production from brief through to live personalised delivery.
All three areas focus on controlled AI deployment: agents can connect to third-party models and tools, but every interaction runs within governed workflows that apply business rules and maintain compliance records. Pega positions this governance layer as the safeguard that makes the flat-fee model commercially viable.
Pega also released an AI Token Cost Calculator alongside the launch, allowing prospective customers to model token cost savings against their current AI spending. The release is available now via Pega Cloud or customer-managed cloud environments.
To stay across the latest in cloud, AI and enterprise tech analysis from Compare the Cloud, subscribe to our weekly newsletter at https://www.comparethecloud.net/newsletter
