Ping An Scales AI Deployment with 300B Daily Tokens, 100M User…

Event summary

Ping An's AI-powered Express Service hit 100M user interactions in 4 months, resolving 92% of requests end-to-end.
AI Doctor now handles 25% of in-hospital services, with 91% diagnostic accuracy for breast/gastric cancer.
Daily token consumption surged from 30B to 300B, supported by 2x computing capacity growth.
AI-generated code accounts for 80% of Ping An's total output, with 20% of tasks fully automated.

The big picture

Ping An's presentation at the 2026 Apsara Conference highlights the growing tension between AI's general capabilities and the need for domain-specific intelligence in highly regulated industries. The company's ability to scale AI deployment while controlling costs positions it as a leader in operational efficiency, particularly in its integrated finance and healthcare strategy. With computing capacity doubling and token consumption increasing tenfold, Ping An is setting a benchmark for large-scale AI implementation in financial services.

What we're watching

Cost Efficiency: Whether Ping An can maintain 76% cost reduction per token as AI deployment scales further.
Domain Specialization: How Ping An's integrated framework for domain-specific intelligence will differentiate it in finance and healthcare.
Operational Integration: The pace at which AI will fully automate more routine development and customer service tasks.