Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.uselamina.ai/llms.txt

Use this file to discover all available pages before exploring further.

The generative media platform for AI agents. You call apps — not models. Lamina handles orchestration, brand context, quality scoring, and multi-channel publishing.

30+ creative apps

Image gen, video, try-on, compositing, upscaling, and more

5 MCP tools

Native agent integration with OAuth

15+ AI clients

Claude, Cursor, Windsurf, VS Code, and more

Choose Your Path

MCP Server

Give your AI agent creative superpowers. Connect via OAuth — no API keys to manage.

REST API

Build generative media into your product. Full control over creation, scoring, and distribution.

CLI

Run creative workflows from your terminal. Ideal for scripting and local development.

Get Started in One Line

npx add-mcp https://app.uselamina.ai/mcp/agent

What You Can Generate

Product ShootsProduct Shoots — Hero shots, lifestyle scenes, catalog-ready assets

Virtual Try-OnVirtual Try-On — Garment-on-model from shopper photos

Campaign BannersCampaign Banners — Multi-format ads and social banners

Instagram ReelsVideo & Reels — Short-form video and product animations

Ad VariantsAd Variants — A/B creative variants at scale

Brand FilmsBrand Films — Cinematic content and storyboards

How It Works

1

Create

Use POST /v1/content/create for intent-driven creation, or POST /v1/apps/{appId}/runs for precise control. Both return a run ID immediately.
2

Track

Receive results via webhook callback, SSE stream, wait endpoint, or poll GET /v1/runs/{runId}.
3

Evaluate

Score outputs with POST /v1/content/score for quality and brand alignment signals.
4

Distribute

Publish via POST /v1/publishing/publish to connected channels, or transfer assets to your CDN.

Built for AI Agents

If you are building with Claude, GPT, LangChain, or any agent framework, Lamina is designed for programmatic consumption from the ground up.
  • Schema-first — Every app has a discoverable, typed input contract. Agents can inspect inputs before calling.
  • Async-native — Executions return immediately. Results arrive via webhook, SSE, or long-poll — no blocking.
  • Intelligence-grounded — Brand context, performance predictions, and trend data are all API-accessible.
  • One-call creationPOST /v1/content/create goes from intent to asset without manual orchestration.
  • MCP-native — Connect as a remote MCP server with OAuth. No key pasting, no tool configuration.
Read The Creative Engine for the full mental model. For machine-readable specs, fetch /llms.txt.

Authentication

All endpoints use a workspace-scoped API key.
x-api-key: lma_your_api_key
or:
Authorization: Bearer lma_your_api_key
MCP connections use OAuth instead — see MCP OAuth Install. Read Authentication for the full model.

Start Building

Quick Start

Create your first asset in under 5 minutes

The Creative Engine

Understand the mental model behind Lamina

Apps and Executions

How apps, runs, and outputs fit together

Agent Integration Patterns

Patterns for building agent workflows with Lamina

Capability Recipes

Common integration patterns and recipes

API Reference

Full endpoint documentation