Gemini 2.5 Pro and Flash are deprecated, auto-route from 1 December
From 2 November both models are gone from Google AI Studio. From 1 December, API requests to them route automatically to newer models, with breaking parameter changes.
Google deprecated Gemini 2.5 Pro and Gemini 2.5 Flash in the Gemini API on 2 November, pulling both from the model picker in Google AI Studio. From 1 December, requests still addressed to gemini-2.5-pro or gemini-2.5-flash route automatically to gemini-3.8-flash and gemini-3.5-flash-lite, billed at those models’ rates.
The swap is not a clean drop-in. Google’s changelog lists breaking changes for anyone who migrates: thinking_level replaces the thinking_budget parameter, some sampling parameters and prefilled model turns are removed, and function calling plus thought_signature validation behave differently. Google is telling developers to update their model string and read each model’s migration guide before the December cutover rather than rely on the auto-route.
gemini-2.5-pro → gemini-3.8-flash
gemini-2.5-flash → gemini-3.5-flash-liteWhy it matters: anything that still hardcodes gemini-2.5-pro or gemini-2.5-flash keeps running past 1 December, but silently switches model, price and behaviour underneath you. Anyone using thinking_budget, the removed sampling parameters or prefilled turns needs to test against the replacement models before then, not after.