picsha 0.2.0

Official Python SDK for Picsha AI - Store, Encode, Understand.

Picsha AI Python SDK 🐍

The official Python client for Picsha AI. Designed specifically for data scientists, computational biologists, and ML engineers.

Installation

We recommend using poetry:

poetry add picsha

Or pip:

pip install picsha

Quickstart

import picsha

client = picsha.Client(api_key="sk_your_key_here")

# 1. Semantic Search
results = client.search(
    query="fluorescent cancer cell cultures", 
    mode="ai"  # Hybrid Vector Search
)

# 2. Get transformed URLs for ML pipelines
for asset in results.assets:
    url = asset.generate_url(width=512, height=512, format="webp")
    print(url)

Asynchronous Support

For high-throughput workloads, an AsyncClient is included out-of-the-box:

import picsha
import asyncio

async def main():
    async with picsha.AsyncClient(api_key="sk_your_key_here") as client:
        result = await client.search(query="cancer cells", mode="ai")
        print(f"Found {result.total_count} assets.")

asyncio.run(main())

See the notebooks/ directory for full examples integrating with Pandas, PyTorch, and SciPy.