As of November 9, 2026, Google says the original Gemini 3 Pro Preview API model was shut down on March 9, 2026. For a version-specific API call, Google lists Gemini 3.8 Flash (gemini-3.8-flash) as stable. The Gemini app offers model choices based on your account, but its general selector labels do not confirm a specific API model ID. (ai.google.dev)
Use Gemini in the app
- Open Gemini or the official mobile app and sign in.
- Choose from the model options shown for your account and submit a prompt.
Google says available models and usage depend on your plan and remaining usage. Its current help page does not establish that a particular app selector label corresponds to the API model gemini-3.8-flash. If you need to specify a model by ID, use the API steps below. (ai.google.dev)
Call Gemini 3.8 Flash through the API
Google lists gemini-3.8-flash as a stable model. The following first-call example uses Google’s Python SDK and Interactions API. (ai.google.dev)
1. Get an API key
Create or view a Gemini API key in Google AI Studio. Google associates each API key with a project; AI Studio creates a project and key for new users. Set the key as an environment variable so genai.Client() can use it. (aistudio.google.com)
Google marks this model as paid. For paid-tier access, link Cloud Billing to your project and prepay at least $5 (or the equivalent in another currency) in credits. (ai.google.dev)
2. Install the SDK and set the key
In a terminal, run:
python -m pip install -U google-genai
export GEMINI_API_KEY="YOUR_API_KEY"
Keep the key server-side or in a secrets manager; do not expose it in public source code or a browser-based production app. (aistudio.google.com)
3. Make a test request
Save this as try_gemini.py, then run python try_gemini.py:
from google import genai
client = genai.Client()
interaction = client.interactions.create(
model="gemini-3.8-flash",
input="Explain how AI works in a few words"
)
print(interaction.output_text)
A successful call prints the response text. This matches Google’s getting-started example for the SDK method, model ID, input field, and output_text accessor. If authentication or model access blocks the request, use Google’s API key setup guide and API troubleshooting guide; Google also documents Interactions API error codes. (aistudio.google.com)
Price and usage limits
Google lists introductory rates of $0.75 per 1 million input tokens and $3.75 per 1 million output tokens through December 31, 2026. Published standard rates take effect January 1, 2027: $1.50 per 1 million input tokens and $7.50 per 1 million output tokens. These are token-based API rates, not a monthly app subscription. For example, 100,000 input tokens and 20,000 output tokens at the introductory rates would cost about $0.15: 0.1 × $0.75 + 0.02 × $3.75. Check Google’s pricing before budgeting for usage after the introductory period. (cloud.google.com)
Rate limits depend on the model and project. Check active limits and usage in Google AI Studio. For limit errors, see Google’s rate-limit guidance. (aistudio.google.com)
Don’t rely on “shadow release” workarounds
An online article claims that choosing “Gemini 2.5 Pro” with Canvas secretly routed requests to Gemini 3.0 Pro. It does not provide official confirmation that this routing was real or remains available. Output style—such as polished code or animations—does not verify which model generated a response. For a version-specific API call, use a model ID listed in Google’s developer documentation rather than an unverified hidden route. (ai.google.dev)