Channels · your people, your tools, your agent
Lamdis is channels, the kind you already work in. The difference is that one member is an agent that has read every message in them, keeps working after you close the laptop, and runs on whichever model you bring.
Here is a channel somebody is in the middle of. Ask the agent about it. Nothing to sign up for.
Free and open source. Nothing to install, nothing to fill in. Mac, Linux and Windows too.
The problem, both halves of it
On one side you have an assistant that is clever and has amnesia. You explain your situation, it helps, and next week you explain it again. On the other you have years of channels holding every decision anyone made, which nobody has time to go back through.
Lamdis is the one place those meet. The channel keeps everything, the way a channel always has. The agent in it has actually read the lot, and answers from what was really said, with the day it was said.
What is in a channel
A channel is a subject: a project, a house move, an incident, a customer. Everything about it goes in one place, and every line says who put it there.
A colleague, a contractor, a client, or nobody at all. You can hand someone the whole channel or only a summary of it, by link, and take it back later. Their agent can be given the same access yours has, so the two of them work from the same facts.
Your issue tracker, your calendar, your own service: paste the address and it joins the channel. Anything that changes something asks you first. Claude Code and Codex read and write the same channels, so two agents stop repeating each other's mistakes.
It is in the channel. It reads what lands, checks it against everything already there, answers when asked and speaks up when it should. What it did is written in the channel like anybody else's message, signed, including what it cost.
While you are not there
In plain words, pinned to the channel: what to do when something new arrives, and what to stand back and think about each morning. From then on it works on its own, and everything it does is posted in the channel for you to read when you get back.
Something lands and it checks it against everything else in the channel, then says what does not add up.
Give it an hour. A morning pass that asks what needs you today; an evening one that looks for what is worth sleeping on.
When the call is yours it asks, in the channel, with options, and waits. It will not commit you to anything and it cannot give anybody access.
Leave one running on your laptop with lamdis agent and direct it from a channel on your phone. It works where you let it, asks before going further, and writes down everything.
Same idea, three different lives
Quotes, dates, the policy number, what the builder said on the phone. In March you write it down. In September you ask which quote was cheaper once the skip hire is counted, and it tells you, and tells you when you wrote each part.
Every price you gave, every date you committed to a customer. Ask before you answer them again. Leave it watching and it raises the clash itself: you told the bank mid-November, and this supplier now starts six weeks out.
Everyone and every agent working the same channel, and afterwards a record you can actually defend: what was read, what was changed, which tools were called, what it cost, who allowed it. Access is granted by a person and can be taken back.
Bring your own key
Everyone else picks the model for you and charges a subscription on top. Lamdis takes your key and talks to whatever you point it at: hundreds of models through OpenRouter, or one running on your own hardware for nothing. Every run in a channel shows what it cost, in cents, beside what it did.
The key sits on your machine and goes to your provider. We are not in the middle of it, we do not mark it up, and there is no subscription to cancel.
Point it at Ollama or vLLM on your own hardware and the whole thing runs locally: your channels, your agent, your model. No key, no bill, nothing leaving the machine.
| Model | Per million tokens, in / out | A typical question |
|---|---|---|
| OpenAI GPT-5.6 Luna default | $0.20 / $1.20 | less than a penny |
| Anthropic Claude Sonnet 5 | $2.00 / $10.00 | a few cents |
| Anthropic Claude Opus 5 | $5.00 / $25.00 | a few cents more |
| Your own hardware | free | free |
Switch between them mid-conversation. No key yet? Write to us and we will send you one with credit on it.
Getting started
Open it and you are in a channel, typing. No account, no card. Add an email later if you want it on your phone too.
Mac and Linux. Windows: irm https://lamdis.ai/install.ps1 | iex
They join your channels and write what they did into them, labelled, like any other member.
Reads the repository, makes the change, runs your tests, and posts what it did.
Straight answers
No, and we are not affiliated with them. Your team chat is where people talk to each other. This is where the work and the decisions live, with an agent in the channel that has read all of it.
ChatGPT forgets between sessions and picks the model for you. A channel keeps everything anyone put in it, runs on whichever model you choose at cost, and its agent can act while you are away.
No. Point Claude Code or Codex at Lamdis and they work in the same channels. It is the memory underneath, not a replacement for the tool you like.
In one file. On your computer, or on ours if you would rather not run anything, and you can take that file and run it yourself whenever you like.
Your provider, and nobody else. It stays on the node you run, we never proxy your requests, and a local model needs no key at all.
It reaches only the tools and sites you allowed, anything that changes something asks first, it refuses credential stores outright, and every single thing it did is in the channel.
Ready?
app.lamdis.ai — or keep the whole thing on your own machine:
Mac, Linux and Windows. Free and open source, on GitHub.