Gemini 2.5 not available in a new project: what to do
Google's September 18 changelog limits Gemini 2.5 models to users who already used them. New projects are pointed to 3.5 Flash-Lite or 3.8 Flash.

If a Gemini 2.5 model is unavailable in a new project, that is expected: the Gemini API changelog entry for September 18, 2026 says access to the 2.5 models is limited to users who have actively used them in the past. Switch the new project to the models Google names for new work, 3.5 Flash-Lite or 3.8 Flash.
The vendor facts below are from the changelog page read 2026-10-01. The Sume side is from Models.
What exactly did Google change?
The entry is titled "Gemini 2.5 models access update". It says the limit exists to keep reliable performance for everyone, and that the models are not deprecated and will continue to be served until further notice through the API. For new projects it says to use 3.5 Flash-Lite or 3.8 Flash, so that capacity stays sufficient for ongoing legacy workflows and new applications.
So this is an access rule, not a shutdown. An existing project that already called a 2.5 model is the case the page describes as continuing; a project created now is the case it redirects.
| Question | What the page says |
|---|---|
| Deprecated? | No, "not deprecated" |
| Who keeps access? | Users who have actively used the 2.5 models |
| New projects | Use 3.5 Flash-Lite or 3.8 Flash |
| Served until | Further notice, through the API |
What should I change in my code?
Treat the model id as configuration, not a literal scattered through call sites. Read it from one environment variable or settings object, so moving a new project to a 3.x model is a one-line change and an old project keeps its current value.
Then re-run your prompts against the new model before shipping. The changelog gives no behavior-equivalence statement, so do not assume output length, tone or structured-output behavior carries over.
How does Sume handle model ids that change?
Sume publishes its own public model ids and marks routes that are on their way out. In its Models page, retiring routes sit under "Deprecated / retiring compatibility aliases — still supported", and the page states that both aliases keep accepting the same body, so a migration there is a URL change only. That example is for the Fabric talking-avatar routes, not for Gemini text models; it shows the pattern Sume documents: a retiring id keeps working while you move.
If part of your pipeline is media generation through Sume, pin the public model id the docs list for that step and re-check the Models page when a vendor changes access. For error handling on the Sume side, see Sume API errors and rate limits.
What if I depend on 2.5 and have no prior usage?
The changelog does not describe an appeal or allowlist process, so do not plan on one. Check the page again for updates, and plan the move to 3.5 Flash-Lite or 3.8 Flash. Related image-model timing is covered in Gemini 2.5 Flash Image shutdown.
Sources
Related posts
More in Developers
- gemini-3.1-flash-tts-preview replacement: 3.8 Flash-Lite TTS
Google names gemini-3.8-flash-lite-tts as the replacement for the 3.1 preview TTS model. What that changes for you, and what stays fixed on Sume TTS.
- Gemini TTS API polling: the Sume audio job pattern
Gemini 3.8 Flash TTS went GA on Sep 22. On Sume, speech is a job: submit async, store the job id, poll the status URL, and never resubmit while it runs.
- Gemini 3.8 TTS returns WAV by default; Sume TTS returns MP3
Gemini 3.8 TTS unary requests return WAV with a RIFF header. Sume TTS 1.0 defaults to mp3 at 44100 Hz, so ask for wav when your pipeline needs it.
- Gemini TTS reads stage directions aloud: Sume emotion field
Gemini 3.8 TTS speaks its text field verbatim, so style goes in speech_metadata. On Sume, keep the transcript clean and put delivery in generation_config.
Written by Sume