Does it need an account?

No. The app runs locally with your own model keys. An account only exists for hosted models and cloud deploy on Pro.

Where does my data live?

On your disk. Context and memory are plain-text files under ~/.letronna. Nothing leaves the machine unless you connect a hosted model.

Which models can I use?

Any OpenAI-compatible endpoint, Anthropic, local models through Ollama or llama.cpp, and the hosted pool on Pro.

Does it keep memory between sessions?

Yes. Memory is written to disk as you work and loaded on the next launch. Edit or delete it like any other file.

What happens when I quit?

Scheduled runs pause unless you have deployed to cloud. Nothing runs on your machine while the app is closed.

The agent runtime is Apache-2.0 on GitHub. The hosted model pool and cloud deploy are the paid parts.

עזרה

Troubleshooting

Checking the setup with letronna doctor, what common errors mean, and where the logs are.

העמוד הזה עדיין לא תורגם, ולכן הוא מוצג באנגלית. אפשר לתרגם אותו ב־GitHub. GitHub ↗

Check the setup

letronna doctor

It checks the version, your system, the data folder and the space left on its disk, the settings, where your keys are kept, the logs, the last crash, and whether the gateway answers. It ends with "No problems found." or says what is wrong. To check without using the network:

letronna doctor --offline

When a message fails

A failed request says why in the chat, and in the window offers the fix. The usual reasons:

It says	What to do
the key was refused	check the key under Settings > Accounts, or with `/connect`
out of credit	add credit on Billing, switch to a gateway with your own key, or set a fallback model
a rate limit	wait the time it names; Letronna waits out short limits by itself
the provider is busy	try again, or pick another model
the conversation is too long	`/compact` summarises earlier turns; `/new` starts fresh

/fallback names models to carry on with when yours is overloaded or out of credit:

/fallback

Signing in

If the browser does not open, the app offers to open the page again or copy its link.
/whoami shows who is signed in on this computer; /login signs in again.
Devices lists the computers signed in to your account.

A model on your own server answers badly

If Letronna does not know a model's limits, set them in the environment:

LETRONNA_CONTEXT_WINDOW=32000 LETRONNA_MAX_OUTPUT=4096 letronna

context_window and max_output in settings.json do the same for good.

Logs and crashes

Letronna keeps a log in logs/ and crash reports in crashes/, in your data folder. They stay on your computer. letronna doctor shows where they are and the last crash.

Starting over

Run setup again to change your name, gateway, model or what is installed:

letronna setup

Uninstalling keeps your data folder, so your sessions and memory stay.

Still stuck

Write to support@letronna.ai with what you did, what it said, and the output of letronna doctor.

עריכת העמוד ב־GitHub ↗

הקודםBuilding on Letronna