Introduction
BetaWe just launched Apps: hosted models, actions and pages, on the same bill as your inference.Read the Apps docs
Docs/Introduction

Inference Bureau docs

Guides and API reference for every Inference Bureau product. Each product has its own section; Platform covers what they share, and the API reference lists every endpoint.

Products

ProductWhat it covers
AppsBuild Shopify and web apps from models, actions and Next.js pages: Inference Bureau stores the data, runs the code, hosts the pages and keeps development and production apart.
AuditorReviews pull requests on GitHub: a rules pass against your repository's conventions and a deep review for bugs, billed as inference. Automations run an agent on a schedule, a pull request or a push across the repositories you choose. For Shopify themes it also deploys theme previews, writes theme documentation and checks merchant content on the store.
GatewayOne inference API, one key and one base URL for every certified model, with Directives, routing, mixed-model sessions, a prepaid balance and a receipt in Logs for every request.

Shared by every product

Every product runs in the same team, with the same members, projects and prepaid balance. Auditor reviews, for example, are billed as inference to that balance, so budgets and Logs cover them too.

SectionWhat it covers
PlatformWhat every product shares: the team, its projects and environments, members and roles, the prepaid balance, and alerts and webhooks.
API referenceEvery endpoint and parameter: inference, routing, spend, the management API and Auditor.

Concepts

TermMeaning
TeamOwns the balance, members, and team-wide Directives.
ProjectGroups environments and keys for one application.
EnvironmentA stage within a project: production or development. Every key belongs to one.

The Gateway overview covers the rest of its terms: Directives, Dispatch and Logs.

Start here