Start here

Quickstart

Create an account, run Lemma locally or in the cloud, install the CLI, and create your first pod.

Choose your path

Use Lemma Cloud when you want a hosted workspace, Lemma Stack when you want the full product on your machine, or the source checkout when you are contributing to the platform.

Which setup should you use?

Path	Use it when	You get
Lemma Cloud	You want to start without hosting anything.	An account at lemma.work and hosted pods.
Local stack	You want the product running on your own machine.	Frontend, API, database, auth, Redis, and document processing in containers.
Source checkout	You want to change Lemma itself.	Hot-reload backend, frontend, and agentbox from this repo.

Cloud quickstart

Create an account at lemma.work, then install the CLI and log in. Browser login will finish account setup if needed.

uv tool install lemma-terminal
lemma servers cloud --use
lemma auth login

Local stack quickstart

Run the released stack on your machine, open the app at http://localhost:3711, create an account, then point the CLI at the local server.

curl -fsSL https://raw.githubusercontent.com/lemma-work/lemma-platform/main/install.sh | bash
uv tool install lemma-terminal
lemma servers select local
lemma auth login

Developer checkout

Use this path only when you are changing the platform source. The dev stack uses 3710/8710 so it can coexist with the installed local stack.

git clone https://github.com/lemma-work/lemma-platform.git
cd lemma-platform
make dev

uv tool install --force --editable lemma-cli
lemma servers add local-dev --base-url http://localhost:8710 --auth-url http://localhost:3710/auth
lemma servers select local-dev
lemma auth login

Create a starter pod

A pod is the workspace boundary: tables, files, agents, workflows, permissions, apps, and surfaces for one team or process.

lemma pod create my-team --with-starter
lemma pods select my-team --save-default
lemma pod describe

Install SDKs for app and function code

Use the TypeScript SDK in frontend apps and the Python SDK in pod function code.

# TypeScript SDK for app frontends
npm install lemma-sdk

# Python SDK for pod function code
uv pip install lemma-sdk

Use the TypeScript SDK

Point the client at a selected pod and start reading tables, running workflows, and chatting with agents.

import { LemmaClient } from "lemma-sdk";

const client = new LemmaClient({ podId: "<pod-id>" });
await client.initialize();

const tables = await client.tables.list();
const runs = await client.workflows.list();

Write pod function code (Python)

Inside a function, the SDK reads credentials from the pod environment automatically.

from lemma_sdk import Pod

pod = Pod.from_env()
rows = pod.records.list("tickets", limit=10).to_dict()["items"]

Next steps

Read the Overview to understand pods, resources, and the mental model.
Follow First Pod to model tables, functions, agents, and workflows.
Browse the CLI section for the full command reference.
Check the SDK section for React hooks, auth, conversations, and data access.