Pymavi

Python SDK for the Mavi Video AI Platform. This SDK provides a simple and intuitive interface to interact with the Mavi API for video processing, search, and analysis.

Installation

pip install pymavi

Quick Start

from pymavi import MaviClient

# Initialize the client
client = MaviClient(api_key="your_api_key")

# Upload a video
response = client.upload_video(
    video_path="path/to/video.mp4",
    callback_uri="https://your-callback-url.com/webhook"
)

# Search for videos
results = client.search_video("find me videos with cars")

# Chat with videos
response = client.chat_with_videos(
    video_nos=["video_id_1", "video_id_2"],
    message="What's happening in these videos?",
    stream=True
)

Features

• Video upload and management
• Natural language video search
• Key clip extraction
• AI-powered video chat
• Comprehensive error handling
• Type hints and documentation

API Reference

MaviClient

The main client class for interacting with the Mavi API.

Methods

• upload_video(video_path: str, callback_uri: Optional[str] = None) -> Dict[str, Any]
• search_video_metadata(start_time: Optional[int] = None, end_time: Optional[int] = None, video_status: str = "PARSE", range_bucket: int = 1, num_results: int = 10) -> Dict[str, Any]
• search_video(search_query: str) -> Dict[str, Any]
• search_key_clip(search_query: str, video_ids: Optional[List[str]] = None) -> Dict[str, Any]
• chat_with_videos(video_nos: List[str], message: str, history: Optional[List[Dict[str, str]]] = None, stream: bool = False) -> Union[str, Dict[str, Any]]
• transcribe_video(video_id: str, transcribe_type: str = "AUDIO", callback_uri: Optional[str] = None) -> str
• get_transcription(taskNo: str) -> Dict[str, Any]:
• delete_video(video_ids: List[str]) -> Dict[str, Any]

Exceptions

• MaviError: Base exception for all Pymavi-related errors
• MaviAuthenticationError: Raised when there are authentication-related errors
• MaviAPIError: Raised when the API returns an error response
• MaviValidationError: Raised when there are validation errors in the input parameters

Development

1. Clone the repository
2. Install development dependencies: pip install -e ".[dev]"
3. Run tests: pytest

License

This project is licensed under the MIT License - see the LICENSE file for details.