Google: Gemini 2.0 models shut down, users move to 3.5 Flash

On 1 June 2026, Google officially retired the Gemini 2.0 Flash line of models from its Gemini API. This is a routine step, but an important one for developers: applications that until then called the older identifiers must switch to the newer models in order to keep working. The change was announced in the official Gemini API changelog.

Which models stopped working?

Four identifiers were taken out of use: gemini-2.0-flash, gemini-2.0-flash-001, gemini-2.0-flash-lite and gemini-2.0-flash-lite-001. “Flash” denotes Google’s fast and economical class of models intended for high request volumes, while the “Lite” variant targets even lower cost and latency. After the shutdown date, calls to those names no longer return a response.

What does Google recommend as a replacement?

Developers are directed to use gemini-3.5-flash or gemini-3.1-flash-lite. The first is a direct successor to the standard Flash line, and the second covers scenarios where low cost and short response time are priorities. In practice, migration usually means replacing the model name in the configuration, but the behavior and pricing of the newer models may differ, so Google advises checking against the specific needs of the application.

Why this matters for developers

Shutting down older models is part of the regular lifecycle of the Gemini platform. For teams that rely on automation tied to a fixed model name in production, such changes require timely code fixes and re-testing. Anyone who does not update their integration risks having a call to the shut-down model simply stop working.