Google retires Gemini 3.7 Flash and 3.5 Flash; API calls go to newer models

On October 8, 2026, Google deprecated two Flash models in the Gemini API and began sending their traffic to newer ones. Requests to gemini-3.7-flash now go to Gemini 3.8 Flash, and requests to gemini-3.5-flash go to Gemini 3.6 Flash. The old model names keep working, but you no longer get the old model.

What changed

Old model ID	Released	Now served by
gemini-3.7-flash	August 13, 2026	gemini-3.8-flash
gemini-3.5-flash	May 19, 2026	gemini-3.6-flash

Google’s deprecations page says the endpoints are not shutting down: the API routes requests to the replacement automatically. Neither old model can be picked in Google AI Studio any more.

What it costs

The deprecations page doesn’t say which price applies to routed requests, and the pricing page no longer lists Gemini 3.7 Flash. Gemini 3.8 Flash and Gemini 3.6 Flash both list $0.75 input and $3.75 output per 1M tokens through December 31, 2026, rising to $1.50 and $7.50 on January 1, 2027.

Other changes in the same weeks

• Deep Research preview: deep-research-pro-preview-12-2025 shuts down on October 23, 2026. Google points users to deep-research-preview-04-2026 or deep-research-max-preview-04-2026.
• Gemini 2.5: since September 18, 2026, the 2.5 models are open only to users who used them recently. Google says they are not deprecated and will be served until further notice, but it recommends Gemini 3.5 Flash-Lite or 3.8 Flash for new projects.

What to do

A routed model can answer differently, use a different number of thinking tokens, or reject a setting the old one took. For example, Gemini 3.8 Flash returns an error for the minimal thinking level. If your app called either old model, test its main prompts again, then change the model ID in your code so the version you run is the one you chose.