Docs

Build on the enterprise AI OS.

Everything you need to integrate, deploy, govern, and scale AI on Deepstack.

Make your first governed, multi-model request in under five minutes.

Route across OpenAI, Claude, Gemini, local, and open-source LLMs with fallback.

Build AI agents, workflow chains, memory, tool execution, and multi-agent systems.

Vector search, semantic retrieval, embeddings, and context engineering.

Trace prompts, latency, hallucinations, failures, tokens, and cost.

Permissions, audit logs, AI policies, compliance, and deployment controls.

Ship to managed cloud, BYOC, VPC, or on-prem with one config.

Typed TypeScript and Python clients for the full Deepstack API.

Complete REST and streaming API reference for every module.