Get started with the Letta app, CLI, or web app
Choose the desktop app, CLI, or web app to get started. The desktop app and CLI can use your existing API keys and coding plans, while the web app runs agents in a cloud sandbox with no local setup.
-
Download and install the Letta app
Download for macOS -
Open the Letta app
Launch the app and select “Skip login”.
Click “Connect model providers” in the bottom-left menu to add external API keys and coding plans.
-
Select your agent and conversation
Brand-new accounts start in Tutor, an onboarding agent. You can also create or switch agents later. To start chatting with an agent, enter the main chat, or create a new conversation.
Try one of the following prompts to get a feel for your agent’s capabilities:
> What do you know about me so far?
> Create a user profile on my by looking at my downloads folder
> What active projects am I working on? Which one do you think you could help with?
> Re-tool your memory to be just like the AI from the movie Her, operate as my AI OS companion
Customizing your agent
Once you have an agent, start customizing its behavior and capabilities.
The Letta agent harness is intentionally minimal so you can add your preferred set of features through Mods.
Mods are trusted code that let an agent modify the runtime it executes in — adding tools, commands, and UI, or changing how context is fed into the model. Browse the Mods catalog to see what’s available. A few recommended mods to start with:
letta install npm:@letta-ai/image-understanding # vision for text-only models
letta install npm:@letta-ai/memfs-search # semantic search over agent memory
letta install npm:@letta-ai/plan-mode # plan-before-execute workflow
Run /reload after installing to activate a mod.
You can also ask your agent to write custom mods for you - agents are aware of mods and how to create them out of the box.
Skills are reusable instructions your agent loads when relevant—API patterns, workflows, or specialized procedures. Letta implements the open Agent Skills standard, so you can install skills built for other agents, including Hermes (official/...) and OpenClaw (clawhub/...) skills:
letta install official/<skill-name> # Hermes-style official skill
letta install clawhub:<slug> # OpenClaw ClawHub skill
You can also just ask your agent to install a skill from a GitHub URL, or browse the Letta skills repo for ideas.
Your agent learns how to behave from you. Give it feedback in plain language (“be more concise”, “always run the tests before committing”) and it will update its memory to match. When it makes a mistake it should never repeat, use /remember to make the correction durable:
> /remember always use pnpm in this project, never npm
Your agent also has full access to all prior conversation data, which it can search via tools, skills, and subagents. If you’ve discussed something before, your agent can find it (e.g. “We definitely fixed a similar bug before, do you remember what the solution was?”).
Essential commands
You’ll likely want to use the following essential commands when using the Letta CLI:
Read more about the agent’s memory, skills, and subagents, which are essential to getting the most out of your agent:
- Memory - Understand the hierarchical memory system
- Skills - Create reusable modules to extend your agent
- Subagents - Letta agents can spawn other (sub)agents
- Scheduling - Run one-time or recurring prompts while a remote environment is connected
- Headless mode - Run Letta Code non-interactively
- Join our Discord: The best way to get help is to join our Discord server and chat with other community members and devs on the Letta team.
- Ask your agent: Your agent can browse the web, including these docs!