Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/googleapis/python-genai/llms.txt

Use this file to discover all available pages before exploring further.

Uploads a file to the Gemini File API, making it available for use in generation requests with large files, videos, or audio.

​
Method Signature

client.files.upload(
    file: Union[str, os.PathLike[str], io.IOBase],
    config: Optional[UploadFileConfigOrDict] = None
) -> File

​
Parameters

​
file
string | PathLike | IOBase
required
The file to upload.Can be:
  • A file path as a string: "/path/to/file.pdf"
  • A PathLike object: Path("document.pdf")
  • An IOBase object (file handle): Must be opened in binary mode ('rb') and be seekable
When using an IOBase object (file handle), ensure:
  • The file is opened in binary mode (not text mode)
  • The file is opened in blocking mode (the default)
  • The stream is seekable (supports seek() operations)
​
config
UploadFileConfig
Optional configuration for the upload.Available options:
  • display_name: Human-readable name for the file
  • mime_type: MIME type (auto-detected if not provided)
  • name: Custom resource name
  • http_options: Custom HTTP request options

​
Returns

​
file
File
A File object containing:
  • name: The resource name (e.g., "files/abc123")
  • uri: The URI to reference in API calls
  • display_name: Human-readable name
  • mime_type: The file’s MIME type
  • size_bytes: File size in bytes
  • create_time: When the file was uploaded
  • update_time: Last update time
  • expiration_time: When the file will be deleted
  • state: Upload state (PROCESSING, ACTIVE, or FAILED)

​
Examples

​
Upload PDF Document

from google import genai

client = genai.Client(api_key='your-api-key')

# Upload file
file = client.files.upload(file='research_paper.pdf')

print(f"Uploaded file: {file.name}")
print(f"URI: {file.uri}")
print(f"Size: {file.size_bytes} bytes")

​
Upload with Display Name

# Upload with custom display name
file = client.files.upload(
    file='document.pdf',
    config={'display_name': 'Q4 Financial Report'}
)

print(f"File: {file.display_name}")
print(f"MIME type: {file.mime_type}")

​
Upload Video File

# Upload video
video_file = client.files.upload(
    file='presentation.mp4',
    config={'display_name': 'Product Demo Video'}
)

print(f"Video uploaded: {video_file.name}")
print(f"State: {video_file.state}")

​
Upload from File Handle

# Upload using file handle
with open('data.json', 'rb') as f:
    file = client.files.upload(
        file=f,
        config={
            'display_name': 'User Data',
            'mime_type': 'application/json'
        }
    )

print(f"Uploaded: {file.name}")

​
Upload and Use in Prompt

from google.genai import types

# Upload file
file = client.files.upload(file='annual_report.pdf')

# Wait for processing if needed
import time
while file.state == 'PROCESSING':
    time.sleep(1)
    file = client.files.get(name=file.name)

# Use in generation
response = client.models.generate_content(
    model='gemini-2.0-flash',
    contents=[
        types.Part(file_data=types.FileData(file_uri=file.uri)),
        'Summarize the key points from this report'
    ]
)

print(response.text)

​
Upload Multiple Files

# Upload multiple documents
file_paths = ['doc1.pdf', 'doc2.pdf', 'doc3.pdf']
files = []

for path in file_paths:
    file = client.files.upload(file=path)
    files.append(file)
    print(f"Uploaded: {file.display_name or file.name}")

# Use all files in one prompt
parts = [
    types.Part(file_data=types.FileData(file_uri=f.uri)) 
    for f in files
]
parts.append('Compare and contrast these documents')

response = client.models.generate_content(
    model='gemini-2.0-flash',
    contents=parts
)
print(response.text)

​
Async Upload

import asyncio
from google import genai

client = genai.Client(api_key='your-api-key')

async def upload_example():
    # Upload file asynchronously
    file = await client.aio.files.upload(file='document.pdf')
    print(f"Uploaded: {file.name}")
    return file

asyncio.run(upload_example())

​
Upload with Custom MIME Type

# Explicitly set MIME type
file = client.files.upload(
    file='data.custom',
    config={'mime_type': 'application/octet-stream'}
)

​
Upload Audio File

# Upload audio for transcription or analysis
audio = client.files.upload(
    file='podcast.mp3',
    config={'display_name': 'Podcast Episode 1'}
)

# Use in prompt
response = client.models.generate_content(
    model='gemini-2.0-flash',
    contents=[
        types.Part(file_data=types.FileData(file_uri=audio.uri)),
        'Transcribe this audio and provide a summary'
    ]
)
print(response.text)

​
File Lifecycle

Uploaded files are automatically deleted after 48 hours. Plan your workflows accordingly or re-upload files as needed.

​
Error Handling

try:
    file = client.files.upload(file='document.pdf')
    print(f"Success: {file.name}")
except FileNotFoundError:
    print("File not found")
except Exception as e:
    print(f"Upload failed: {e}")

​
API Availability

This method is only available in the Gemini API (not Vertex AI).For Vertex AI, you should:
  • Store files in Google Cloud Storage (GCS)
  • Use files.register_files() to register GCS URIs
  • Or use direct GCS URIs in prompts

Build docs developers (and LLMs) love

Get started for freeTalk to us