Cloud Next ‘26: Momentum and Innovation at Google Scale

Why It Matters

The moves cement Google Cloud’s leadership in enterprise AI, giving customers faster, more secure, and cost‑effective compute while showcasing a scalable model for AI‑augmented development across the industry.

Key Takeaways

• 16 billion tokens per minute processed, up from 10 bn last quarter
• Gemini Enterprise agents grew 40% MoM in paid monthly active users
• New TPU 8t and TPU 8i deliver up to 3× performance per watt
• AI‑APP with Wiz offers autonomous protection across multicloud environments
• 75% of Google’s new code now AI‑generated, cutting migration time sixfold

Pulse Analysis

Google Cloud’s latest performance metrics underscore a rapid acceleration in AI adoption. By processing 16 billion tokens per minute—an increase of 60% from the prior quarter—the platform demonstrates the scalability needed for enterprise workloads. The shift of more than 50% of the company’s 2026 ML compute budget toward cloud services signals a strategic focus on delivering AI capabilities directly to customers, positioning Google as a primary infrastructure provider for data‑intensive applications.

The rollout of the Gemini Enterprise Agent Platform marks a pivotal evolution from isolated AI models to orchestrated agent ecosystems. With a 40% rise in paid monthly active users, the platform is proving its value in enabling businesses to embed AI agents across operations, from customer service to internal workflow automation. Coupled with the new AI‑powered security suite built with Wiz, Google offers autonomous threat detection and mitigation that reduces incident response times by over 90%, addressing growing concerns about AI‑related vulnerabilities.

Hardware innovations round out the announcement, as Google introduces eighth‑generation TPUs—TPU 8t for massive training workloads and TPU 8i for low‑latency inference. Both chips deliver up to three times the performance per watt of previous generations, delivering cost efficiencies for large‑scale AI deployments. Internally, the company’s reliance on AI‑generated code—now 75% of new code—has slashed development cycles, with complex migrations completed six times faster. These advances not only boost Google’s own productivity but also provide customers with a proven, high‑performance foundation for building the next wave of AI‑driven services.