- Gemini 2.5 Pro and Flash rerouted
Gemini 2.5 Pro and Flash get rerouted on December 1, 2026
Google's Gemini API deprecations page lists December 1, 2026 for gemini-2.5-pro and gemini-2.5-flash. The requests won't fail. From that day the API sends gemini-2.5-pro requests to gemini-3.8-flash and gemini-2.5-flash requests to gemini-3.5-flash-lite, billed at the new model's rates. Both also leave the model list in Google AI Studio.
gemini-2.5-flash-lite is not deprecated and stays available until further notice. Google calls the listed dates the earliest possible ones and says exact dates are announced in advance.
What breaks
Nothing throws, so your error logs stay quiet. A pipeline tuned on 2.5 Pro starts getting a Flash model's answers, and a 2.5 Flash workload moves to a Flash-Lite model. Prompts, output length limits and quality checks written for the old model may not hold.
How to fix it
Pick the next model yourself instead of letting the routing pick it. Change the ID in code and config, rerun your evaluation or a sample of real prompts, and work out the cost at the new model's prices. The scanner finds the 2.5 Pro and Flash IDs, preview IDs included, and leaves -flash-lite alone.
Check a repository for this
gemini-2.5-pro and gemini-2.5-flash model IDs (and their previews) in code and config, not -flash-lite.
These rules live only on this site; there is no command-line version.
Sources
Dates and wording re-read on the official pages on . Vendors move dates; the official page wins.