Google Deprecates Gemini 2.5 Pro and 2.5 Flash: API Will Automatically Reroute Calls Starting December 1

On November 2, 2026, Google updated the Gemini API changelog to deprecate two previous-generation models originally released in June 2025: gemini-2.5-pro and gemini-2.5-flash. Both have vanished from the Google AI Studio model picker, and a deadline is set: starting December 1, 2026, requests targeting them will not be rejected outright, but instead served by a different model without requiring client-side code changes.

What changes on December 1

Requested identifier	Served starting December 1 by	Billed at the rate of
gemini-2.5-pro	gemini-3.8-flash	Gemini 3.8 Flash
gemini-2.5-flash	gemini-3.5-flash-lite	Gemini 3.5 Flash-Lite
gemini-2.5-flash-lite	unchanged (no shutdown announced)	unchanged

Google advises updating model identifiers before this date and points developers to its migration guides. According to the deprecations page, API access to these 2.5 models is already restricted to developers who have used them in the past.

The trap: same name, different model

Google’s approach is unusual. Typically, a retired model returns an error, which immediately alerts developers. Here, requests will continue to return responses, but powered by a different model from an entirely different tier: a customer calling for a Pro model will instead get a Flash model.

Crucially, Google warns that the shift involves breaking API changes:

• stricter validation for function calling and the thought_signature field;
• thinking_level replacing thinking_budget to configure model reasoning;
• removal of sampling parameters and prefilled model turns (“prefill”).

Any application tuning sampling parameters, forcing a response prefix, or relying on 2.5 model function calling risks unpredictable behavioral changes on December 1, even if API calls continue to succeed.

What about the bill?

According to Google’s published rates (paid tier, per million tokens), the 2.5 Pro redirection is initially cheaper:

Model	Input	Output
gemini-2.5-pro (prompts up to 200,000 tokens)	$1.25	$10
gemini-3.8-flash through December 31, 2026	$0.75	$3.75
gemini-3.8-flash starting January 1, 2027	$1.50	$7.50
gemini-2.5-flash (text, image, video)	$0.30	$2.50
gemini-3.5-flash-lite	$0.30	$2.50

Pricing published by Google on November 3, 2026.

The discounted pricing on 3.8 Flash lasts only one month after the cutover: starting in January, input will be more expensive than 2.5 Pro for short prompts, while output will remain slightly lower. For 2.5 Flash, text pricing remains identical, but the replacement model is a tier lighter.