Globe Telecom is pushing artificial intelligence (AI) into the mainstream with AI Fiesta plans starting at P49, betting that lower-priced access will encourage more Filipinos to adopt premium tools.

The telco said Friday the service brings together multiple leading AI models in a single app to allow users to compare responses and choose the best model for different tasks without maintaining multiple subscriptions.

"AI is becoming part of everyday life, especially for Filipinos looking for smarter and more efficient ways to manage work, studies, passion projects, and personal responsibilities," Globe chief marketing officer KD Dizon said.

"Through these Globe-exclusive offers for AI Fiesta, we are making premium AI tools more accessible and relevant to everyday needs, while helping customers discover practical ways AI can support productivity and creativity."

The platform brings together AI models, including ChatGPT, Claude, Gemini, Grok, and DeepSeek, enabling users to generate responses from multiple models using a single prompt.

It also offers features such as Image Studio for content creation, Super Fiesta Mode for intelligent model selection, and tools for deep research and real-time web searches.

Globe said bundling multiple premium AI services into a single app with token-based pricing addresses one of the biggest barriers to AI adoption, which is the cost of maintaining separate subscriptions across different platforms.

"People are increasingly curious about AI, but navigating different tools and subscriptions can be overwhelming," said Mohammad Hasan, chief executive officer and co-founder of AI Fiesta.

"AI Fiesta was designed to simplify that experience by bringing together access to leading AI models, including ChatGPT, Claude, Gemini, Grok, DeepSeek, and others, in one application. This allows users to compare responses, discover which models best suit different tasks, and make AI more useful for learning, work, creativity, and everyday productivity."

Customers can buy token-based access packs, with tokens consumed depending on the complexity of each prompt.