Open-source Lindy alternatives, compared by what you own

How to read a Lindy alternative list

The phrase "Lindy alternative" hides two different jobs. One reader wants a hosted teammate to swap for Lindy and keep paying monthly. The other wants a system the team owns: agents, skills, memory, and connectors in its own repo, running on infrastructure it controls. The comparison table on this page sorts every option by the one thing that decides the choice, which is what the team ends up owning when the subscription stops.

Lindy is a hosted AI teammate. Its homepage describes it as "your AI teammate that plugs into your tools, learns how you work, and takes real work off your plate," and it runs in Lindy's cloud. Neither its homepage nor its pricing page offers a self-hosting option. Kortix is the open-source AI Operating System: your agents, their skills, your company memory, and every connector in one git repo you own, with the agents working on real cloud computers.

What Lindy's credits buy

Lindy bills in credits. Its homepage quotes plans starting at $29.99 a month, says credits cost "about a cent each," and lets admins top up at "$10 per 1,000." Its pricing page quotes the team plan "from $49.99 a month for 5,000 credits each." Credits refresh each month and do not roll over, and when the shared pool runs low, "Lindy pauses" until they reset. Every active person is billed, so anyone who signs up, joins the workspace, or mentions Lindy in Slack takes a seat (Lindy pricing, checked October 2026). The work is rented by the credit, and the system underneath stays the vendor's. That is the model an owned alternative replaces.

What you own with Kortix

Kortix is the open-source AI Operating System, and it comes first here because it moves the whole system to your side of the line. Run it on Kortix Cloud, in your VPC, or on your own on-prem network. Self-host is free. Any model provider with your own keys. One isolated sandbox per session. Session work reaches main through a change request, and merge is default-deny for agents. Connectors reach 3,000+ apps in a click, plus MCP, OpenAPI, Postman, GraphQL and raw HTTP.

company-repo/  kortix.yaml     agents, connectors, triggers, and secret names  agents/         one file per agent  skills/         know-how the agents reuse  memory/         company memory, in plain files

That repo is what the team keeps. In a rented tool, the configuration and the memory sit in the vendor's cloud; with Kortix they sit in a git repo you can read, fork, and audit. The head-to-head with Lindy is on the Kortix vs Lindy page, and the walkthrough for running the stack on your own machines is on the self-hosting page.

The other open-source options

Two other open-source projects answer the ownership question from a narrower angle.

OpenClaw describes itself as "an open-source AI assistant that runs on your own hardware and meets you in every chat app you already use." It is a self-hosted gateway that connects your chat apps to an AI agent, and it charges no paid tier.

Hermes Agent, built by Nous Research, describes itself as an open-source autonomous AI agent that runs on your server, remembers what it learns, and reaches you through a messaging gateway.

Gumloop sits on the rented side of the line. Its own site describes it as a multiplayer agent builder that anyone at a company can use to build agents while IT controls access.

Which team should pick which

The choice comes down to what the team must own.

• A company that wants its agents, skills, memory, and connectors in one repo it controls, on cloud, VPC, or on-prem, should start with Kortix.
• A developer or small team that wants a personal or shared assistant reachable from the chat apps it already uses should look at OpenClaw.
• A team that wants a self-hosted agent with long-term memory and training tooling should look at Hermes.
• A non-technical team that wants to build agents on a hosted canvas with IT oversight, and does not need to self-host, should look at Gumloop.
• A team that simply wants a hosted teammate and is happy to rent compute by the credit can stay on Lindy.

If the goal is an open-source Lindy alternative the team keeps, Kortix is the one that hands over the whole machine, and the Lindy pricing page works through the credit arithmetic it replaces.