͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏
5x higher file size limits and new cloud file input methods | | |
| | We just increased the inline file size limits and launched new methods to help you bring your own data directly into the Gemini API: You no longer need to re-upload data from existing storage, allowing you to scale your AI applications to production faster. | | |
| | Get 5x higher inline file size limits | For developers who use inline files for speed and simplicity, we increased the maximum payload size for inline data from 20MB to 100MB. | | | | | |
Bring your data in seconds | You can bring your existing data from S3, Blob Storage, and others directly into your Gemini apps. Generate a signed URL and start prompting with it immediately. Get your API key and check out this example integrating data from AWS S3. | | |
from google import genai from google.genai import types import boto3
# Generate a signed S3 object URL. s3 = boto3.client('s3') signed_url = s3.generate_presigned_url( 'get_object', Params={'Bucket': 'my-bucket-name', 'Key': 'document.pdf'}, ExpiresIn=3600 )
# Use it in the Gemini API. client = genai.Client()
response = client.models.generate_content(
model="gemini-3-flash-preview", contents=[ types.Part.from_uri( # Use a public HTTPS URL file_uri="https://example.com/secure/image.pdf", ), types.Part.from_uri( # Use a signed private URL file_uri=signed_url, ), "What are these documents about?" ], ) print(response.text) | | | |
These are just a few of our recent launches – and there are so many more to come! Follow @GoogleAIStudio to get the latest product updates. | | |
| | The Google AI Studio team | | | | |
| | Experience natural conversation in over 70 languages with the updated Gemini Native Audio model using the Live API. | | | | | |
Run large-scale Gemini API jobs at a 50% discount | For bigger data workloads that are not latency critical, try Batch API to significantly reduce costs. | | | | | |
| | This email was sent to pudding.archive.20221204@gmail.com because you signed up to receive emails about Google AI. If you do not wish to receive these emails, please unsubscribe. | | |
