Developer tools

Last updated: October 11, 2026

Layer's models, workspaces and Creative Unit balance are also reachable from code, from AI assistants and from the terminal. This page says what each developer tool is for and where its documentation lives; the setup steps are in the developer docs at layer.ai/docs.

How it works

REST API and SDK

Generate images, video, 3D and audio from your own services, build scripts and pipelines. The API is versioned and documented at layer.ai/docs; start with the Quickstart. The official JavaScript and TypeScript SDK wraps the same API.

Authentication

Requests are signed with a Personal Access Token you create in your Layer settings. See Authentication.

Files

Reference images, masks and other inputs are uploaded once and then passed to generations by id. See Files & uploads.

MCP servers

Connect Layer to an AI assistant that supports the Model Context Protocol, such as Claude, Cursor, VS Code or Codex. The creative server generates assets from a conversation (MCP server); a separate management server administers workspaces: projects, groups, members and usage (Management MCP server). In the app, Connect MCP at the bottom of the sidebar opens a page with the setup steps for each assistant.

CLI

The layer command runs generations from a terminal, a Makefile or a CI job. See CLI.

What it means for you

  • These tools reach the same models, workspaces and Creative Unit balance as the app, so what they generate is spent from the workspace you point them at.
  • A Personal Access Token is shown only once, when you create it. Keep it in a secret manager, never in source control, and create a new one if it is lost.
  • Developer questions, changes and deprecations are covered in the developer docs, not this help center. API changes are listed at layer.ai/docs/changelog.

Related