Haystack / deepset
Haystack is deepset's open-source Python framework for building context-engineered, production-ready LLM applications - modular Pipelines and agent workflows assembled from 100+ Components and document stores for RAG, semantic search, and agents. Haystack pipelines deploy as REST services via Hayhooks, and the commercial deepset AI Platform (deepset Cloud) exposes a hosted REST API at api.cloud.deepset.ai for pipelines, search, files, and workspaces.
APIs
Haystack Framework (Open Source)
Apache-2.0 licensed Python framework for composable LLM orchestration. Build modular Pipelines from Components (retrievers, routers, embedders, generators, evaluators) over Docu...
Hayhooks REST Deployment
Hayhooks turns a Haystack pipeline into a self-hosted REST API (and optional MCP server) with one command, auto-generating OpenAPI/Swagger docs and HTTP run endpoints. The deplo...
deepset Cloud API - Pipelines
Hosted REST API to create, list, deploy, and undeploy Haystack pipelines inside a deepset Cloud workspace. Bearer-authenticated under /api/v1/workspaces/{workspace}/pipelines.
deepset Cloud API - Search
Runs queries against a deployed pipeline, returning answers and retrieved documents. POST /api/v1/workspaces/{workspace}/pipelines/{pipeline}/search with queries, filters, and r...
deepset Cloud API - Files
Uploads, lists, and deletes files in a workspace, updates file metadata, and manages chunked upload sessions for indexing into pipelines and indexes.
deepset Cloud API - Workspaces
Creates, lists, and deletes workspaces - the isolation boundary that holds a deepset Cloud account's pipelines and data. Pipelines and data are not shared across workspaces.