Lower inference costs from Frozen v2 will enable Google to expand free user offerings, offer competitive Cloud pricing, and improve model training through increased interaction data
Too little corroboration in the last 3 days to call a trend (3 articles). Watching for it to gain traction.
Lower inference costs from optimized model versions like Frozen v2 will allow Google to expand free-tier offerings to users while maintaining competitive Cloud pricing and generating more training data through increased user interactions. This cost reduction addresses Google's current gap in frontier models by shifting strategy toward efficiency and data accumulation rather than raw capability.
Inference cost curves directly influence the unit economics of AI services and determine which business models become viable at scale. When costs fall, companies can afford to offer free or freemium tiers that drive user volume and data collection, creating a flywheel that competitors with higher cost structures cannot match.
"With 3.5 Pro still missing and Anthropic and OpenAI sprinting ahead in the AI race, Google no longer has a frontier model. Instead, it has doubled down on its 'Flash' models, which offer greater speed and lower costs for users, at a time when customers are keeping a closer eye on their AI bill."
"College students can also opt for a Google AI Pro and YouTube Premium bundle for $8.99 per month in the United States, which is a 70 percent discount off the regular price. After the free subscription period, Google will charge $19.99 per month for AI Pro unless the plan is canceled."
"Lower inference costs would also let Google serve more free users, offer more competitive Cloud pricing, and collect more interaction data for model improvement... Frozen v2 could also ease the capacity shortage that reportedly caused Google Cloud to turn away some business."