Quick Reference

A summary of all endpoints available in the deepset API and their purpose.

Collection Overview

The base URL for all endpoints, except for health is https://api.cloud.deepset.ai/api/v1.

Deepset AI Platform

These are the endpoint collections available in this section:

Collection

Purpose

health

Checking the connection.

haystack

Obtaining Haystack components available in deepset AI Platform, including their input and output types.

infrastructure

Managing the connection to your VPC.

user

Listing and deleting users.
Updating user roles.

models

Listing models available for a component.

model_registry_token

Working with tokens for private models.

organization

Listing deepset organizations and adding users to them.

pipeline_utils

Updating pipeline YAML by passing updated data as a dictionary and merging it with the original YAML configuration.

prompts

Working with prompt templates in PromptHub.

api_token

Managing API keys for connecting deepset to your app.

workspace

Working with deepset workspaces. This includes, creating, listing, and viewing workspace statistics and search history.

datasets

Listing the sample datasets deepset offers if you have no files in your workspaces and adding files to them.

file

Working with files in deepset AI Platform. This includes creating, deleting, listing, uploading, and working with metadata.

document_store

Listing indexes, and working with documents: counting, querying, and listing.

pipeline_templates

Listing pipeline templates.

pipeline

Interacting with deepset pipelines. This includes creating, validating, deploying, listing, and getting all pipeline-related information, searching with a pipeline, and more.

search_session

Listing and creating search sessions. Search sessions are used for chat.

shared_prototype

Managing shared prototypes of your pipelines.

upload_session

Opening, closing, and checking the status of the upload sessions used to upload files with the SDK. For more information, see Tutorial: Uploading Files with Python Methods.

secrets

Managing deepset secrets. Secrets are used to connect to service providers through custom components.

debug_tunnel

Starting a debug tunnel for remote debugging with VS Code.

Jobs

These are the endpoint collections related to working with jobs in deepset:

CollectionPurpose
jobsWorking with jobs. This includes creating, listing, updating, starting, and getting results.
shared_jobManaging jobs shared with others.
query_setsManaging query sets used for jobs.

Feedback

These are endpoint collections related to the feedback feature:

CollectionPurpose
feedbackManaging feedback items, which includes getting all feedback, creating, updating, and deleting feedback items.
tagsListing, creating, and deleting feedback tags.

Custom Components

These are the endpoints you can use to interact with custom components in deepset:

CollectionPurpose
custom_componentsListing and importing custom components, and getting their logs.