We are about to deploy an application using gemini flash 2.0 due to attractive costs.
After a break in the project we just found out it is going to be discontinued in 2 months. How does this work for businesses? Are we forced to quit or move to higher (many times more expensive!) model just like that? Last option is to find another provider with similar pricing and quality.
For context, we use the vertexai.generative_models library in python specifically so not default gemini api.
Any insight appreciated.
Hi Marián,
You definitely do not need to move to the expensive “Pro” model or switch providers. This is just standard lifecycle management for these models.
How it works for businesses: Google (like OpenAI) deprecates specific versions (e.g., v2.0) after a certain period, but they always have a direct replacement in the same “tier.” The gemini-2.0-flash The model is being replaced by the newer Gemini 2.5 Flash (or the latest stable Flash version).
The Solution: You simply need to update the model name string in your code. The newer “Flash” versions are designed to keep the same low-cost/high-speed profile as the previous ones.
Since you are using vertexai.generative_models, the fix is just a one-line change:
Python Code :
from vertexai.generative_models import GenerativeModel
# OLD (Discontinuing):
# model = GenerativeModel(“gemini-2.0-flash”)
# NEW (Upgrade to the latest Flash version):
# This keeps your costs low while giving you the latest updates. model = GenerativeModel(“gemini-2.5-flash”)
Summary:
-
Don’t panic: You are not forced into the “Pro” pricing tier.
-
Action: Just bump the version number in your string to the latest available “Flash” model (2.5 or similar).
-
Pricing: It will remain in the attractive low-cost bracket you originally chose.
Hope this helps!
1 Like
Thanks for the reply. But I have to ask - is your response ai generated?
I haven’t noticed price going down for 2.0 model family when 2.5 became GA which makes me wonder if what you describe is really industry standard