Gemini 3.2 Flash spotted in iOS app and AI Studio

Summary

Gemini 3.2 Flash appeared in the official iOS Gemini app and Google AI Studio metadata on May 5, ahead of any official announcement. Leaked pricing: $0.25 per million input tokens, $2.00 per million output tokens — cheaper than Gemini 3 Flash ($0.50/$3.00) on output. Reported performance near or exceeding Gemini 3.1 Pro on coding tasks. Google has not officially confirmed the model; Google I/O (May 19-20) is the likely reveal window.

Implications

• If confirmed, the most cost-efficient model in Google’s lineup — positioned between Flash-Lite ($0.25/$1.00) and Flash ($0.50/$3.00) pricing with near-Pro performance
• A quiet pre-I/O drop suggests Google is staging capability reveals ahead of the conference
• Feeds the token economics thread: downward pricing pressure continues. Flash-tier models at Pro-level performance compress the entire pricing hierarchy
• Feeds the Gemini CLI thread: if 3.2 Flash matches 3.1 Pro on coding, Gemini CLI gets a major cost reduction for agent workloads