Google deprecates Gemini 2.5 Pro and 2.5 Flash, with forced routing from December 1

Archive entry: this covers an event from November 2, 2026, and we published it on November 8, 2026. The details come from sources dated at the time and may have changed since.

On November 2, 2026, Google deprecated gemini-2.5-pro and gemini-2.5-flash, as recorded in the Gemini API changelog. Both models could no longer be selected in Google AI Studio.

What changed

• Google said that, effective December 1, 2026, all API requests to gemini-2.5-pro would route automatically to gemini-3.8-flash and be billed at Gemini 3.8 Flash rates.
• From the same date, all API requests to gemini-2.5-flash would route automatically to gemini-3.5-flash-lite and be billed at Gemini 3.5 Flash-Lite rates.
• Breaking API changes included function calling and thought_signature validation.
• thinking_level replaced thinking_budget.
• Sampling parameters and prefilled model turns were removed.

What it meant for people building with AI

Apps that used Gemini 2.5 would switch to different models and run into breaking API changes unless they migrated before December 1. Because requests were billed at the replacement models' rates, the routing could also change what those apps cost to run, so teams had reason to update their model strings and test the changed parameters before that date.

Migration deadline

Google asked developers to update their model strings before December 1, 2026. The changelog entry was dated November 2, 2026, which left 29 days between the announcement and the start of forced routing.

Sources

• Gemini API changelog: November 2, 2026 (Google AI Studio (Gemini API changelog), November 2, 2026, official)

Written with AI assistance, strictly from the sources listed above. Before it went live, automated checks verified its dates, sources and every language edition.