TouAI Python SDK
Use one Python client to work with Context Layer, Unstructured Data, Deep Research, Data Search, Storage, Billing, Organizations, and Auth.
TouAI Python SDK is the official Python client for the TouAI platform. It gives application developers, agents, and backend services a single entry point for the platform's data, research, storage, and billing APIs.
Package: touai · Version: 0.2.0 · Python: >= 3.10
Why this SDK
- One client for the full TouAI platform
- First-class Python examples for synchronous and streaming workflows
- Built-in retry handling for transient failures
- Typed response models powered by Pydantic v2
Service Map
Context Layer
Connect external data sources, configure sync, and run RAG search through one client.
Unstructured Data
Process documents and videos, manage files, and enable semantic retrieval.
Deep Research
Run non-streaming or streaming research workflows with tool-aware orchestration.
Data Search
Scrape pages, run web search, and launch deep crawls with auto-polling helpers.
Storage
Store files in Data Lakehouse with simple APIs and organization-scoped storage primitives.
Billing and Platform
Manage credits, subscriptions, organizations, and authentication flows.
Installation at a Glance
pip install touaifrom touai import TouAI
with TouAI() as client:
results = client.context_hub.search("quarterly revenue")
print(results.total_results)Client Surface
The SDK exposes platform services as namespaced properties on the TouAI client:
| Property | Service | Description |
|---|---|---|
client.auth | Auth | Register, login, OAuth, profile, context switching |
client.organizations | Organizations | CRUD, members, invitations, permissions |
client.api_keys | API Keys | Create, rotate, revoke, validate |
client.context_hub | Context Layer | 50+ data connectors, sync, RAG search |
client.unstructured | Unstructured Data | Document and video processing, file management |
client.deep_research | Deep Research | Research workflows with streaming SSE |
client.data_search | Data Search | Web scraping, search, deep crawl |
client.storage | Data Lakehouse | Object storage with project ACLs |
client.billing | Billing | Credits, plans, subscriptions, payments |
Recommended Reading Path
Install and authenticate
Start with installation, authentication, configuration, and client lifecycle so every example later in the docs feels familiar.
Pick a core service
Move into Context Layer, Unstructured Data, Deep Research, Data Search, or Storage based on the workflow you are building.
Finish with platform reference
Use the Billing, Organizations, Auth, Error Handling, and Type Reference pages as reference material while integrating.