Vertex AI
gemini-2.5-pro model retirement
Stopped workingSubscribe to Vertex AIShare
Model gemini-2.5-pro reached its retirement date on October 20, 2026. Calls to model gemini-2.5-pro fail since that day.
The error you get
What Vertex AI answers a call to gemini-2.5-pro once the model is gone, as recorded from real responses:
{"error":{"code":404,"message":"Publisher model `projects/{project}/locations/{region}/publishers/google/models/gemini-2.5-pro` was not found or your project does not have access to it. Ensure you are using a valid model name and that the model is available in the specified region.","status":"NOT_FOUND"}}What to use instead of gemini-2.5-pro
Move to gemini-3.8-flash or gemini-3.5-flash, as recommended by Google.
In Vertex AI's words
Model ID: gemini-2.5-pro; Release date: June 17, 2025; Retirement date: October 20, 2026; Replacement model ID: gemini-3.8-flash or gemini-3.5-flash
Facts
- API
- Vertex AI
aiplatform.googleapis.com - Model
gemini-2.5-pro- Released
- June 17, 2025 (models.dev)
- Kind
- Model retirement
- Takes effect
- October 20, 2026
- Calls after that day
- Fail
- Source
- Deprecations page
- Last updated
- Verified
- , against Vertex AI's own pages
Questions
When does gemini-2.5-pro shut down?
Calls to the model gemini-2.5-pro on Vertex AI stopped working on October 20, 2026.
What happens to calls after October 20, 2026?
They have failed since October 20, 2026.
What error does gemini-2.5-pro return once it is shut down?
Vertex AI answers HTTP 404: "Publisher model `projects/{project}/locations/{region}/publishers/google/models/gemini-2.5-pro` was not found or your project does not have access to it. Ensure you are using a valid model name and that the model is available in the specified region.". Send another model to fix it.
When was gemini-2.5-pro released?
gemini-2.5-pro was released on June 17, 2025, according to models.dev.
How do I find the calls this hits in my code?
Okapi reads your repository, finds the calls this change hits and opens a pull request with the fix. It never commits to your branches: every change comes as a pull request you review.
gemini-2.5-pro on other APIs 3
-
Lovable AI Gateway
google/gemini-2.5-proModel retirement Redirected -
OpenRouter
google/gemini-2.5-proModel retirement Stopped working -
Gemini API
gemini-2.5-proModel retirement Redirectsin 25 days
More from Vertex AI 10
-
gemini-2.5-flashModel retirement Stopped working -
gemini-2.5-flash-liteModel retirement Stopped working -
deepseek-ocr-maasModel retirement Stopped working -
deepseek-r1-0528-maasModel retirement Stopped working -
deepseek-v3.1-maasModel retirement Stopped working -
deepseek-v3.2-maasModel retirement Stopped working -
glm-4.7-maasModel retirement Stopped working -
glm-5-maasModel retirement Stopped working -
gpt-oss-20b-maasModel retirement Stopped working -
kimi-k2-thinking-maasModel retirement Stopped working
You fixed this one by hand.
Okapi opens the pull request with the fix before the next breaking change reaches your code.