SynapsAI Python SDK

The official SynapsAI Cloud AI inference Python SDK. Seamlessly integrate multi-modal AI capabilities into your applications.

Installation

Install the latest version of the SDK:

pip install --upgrade synapsai-python

Requirements

• Python >= 3.8
• httpx >= 0.23.0, < 1.0.0
• pydantic >= 2.0, < 3
• typing-extensions >= 4.5, < 5
• Pillow >= 9.5.0, < 11
• numpy >= 1.21.0, < 2

Quick Start

1. Setup Environment Variables

You can configure the SDK using environment variables instead of hardcoding your credentials:

• SYNAPSAI_API_KEY - Your SynapsAI API key
• SYNAPSAI_API_BASE - Custom base URL (Defaults to https://api.synapsai.cloud/v1)

2. Initialization & Example Usage

Here is how to initialize the client and fetch a list of all deployed models on your account:

import os
from synapsai import SynapsAI

# Initialize the client (picks up env variables automatically if arguments are omitted)
client = SynapsAI(
    api_key=os.environ.get("SYNAPSAI_API_KEY"),
)

# Fetch and view deployed models
models_page = client.models.list()
for model in models_page.data:
    print(f"Model ID: {model.id} | Status: {model.status}")

Documentation

Local guides live in docs/ (getting started, every resource, CLI, async, errors).

Online examples: docs.synapsai.cloud.

Available Resources

The SDK provides native support for the following SynapsAI Cloud capabilities:

CLI

After install, the synapsai command is available:

# Upload a model directory into an existing artifact
synapsai upload-model ./my-model --artifact-id artifact-demo-abc12345

# Upload files or directories into a vector store
synapsai upload-vector-store-files vs_abc123 ./docs ./readme.md

Environment variables:

• SYNAPSAI_API_KEY — API key
• SYNAPSAI_API_BASE — API base URL (default https://api.synapsai.cloud/v1)

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.