Lobu

What is Lobu?

Lobu is an open-source backend for running AI teammates that watch, remember and act across a team. It began as a Slack bot and now wraps the OpenClaw agent runtime in the multi-tenant pieces a team needs: each user or channel gets an isolated, sandboxed worker, idle workers shut down to zero and start again when the next message arrives, and a gateway keeps real API keys away from the agents by swapping placeholder tokens at request time. Outbound network access is denied by default and controlled with domain allowlists, and MCP calls are proxied so each person signs in once with OAuth. Agents connect to Slack, Telegram, WhatsApp, Discord, Teams, Google Chat and a REST API, and skills, providers and Nix packages are set from a settings page. Self-hosting is free and runs as one Node process with Postgres embedded, or you can use the managed Serverless OpenClaw service at $0.000463 / vCPU / minute, $0.000231 / GB / minute for memory and $0.15 / GB / month for storage.

Use cases

• Run AI agents for a whole team in Slack
• Give each user a sandboxed agent
• Connect agents to Telegram and WhatsApp
• Keep API keys out of agent workers
• Proxy MCP tools with per-user sign-in
• Host agents without managing servers
• Embed an agent backend in a Node app

Who it's for

• Agent developers
• Platform engineers
• Teams self-hosting AI
• Startups building AI assistants

Why use Lobu

• Agents run in separate sandboxes, so one user's broken environment does not affect anyone else.
• Real API keys stay encrypted in the gateway, so a compromised worker has no secrets to leak.
• Scale-to-zero workers mean idle agents cost $0 on the managed service, with billing by the second while active.
• You can start free in Slack with your own keys, or self-host and keep data on your infrastructure.
• The same open-source stack runs in the managed cloud and self-hosted, so you can migrate without lock-in.

Lobu features

• Per-user and per-channel sandboxed workers
• Serverless workers that scale to zero when idle
• Gateway proxy so workers never receive real credentials
• Domain allowlists for outbound network access
• MCP proxy with per-user OAuth
• Chat platforms: Slack, Telegram, WhatsApp, Discord, Teams and Google Chat
• REST API for webhooks and custom clients
• Skills, model providers and Nix packages set from a settings page
• Agent config synced from GitHub with lobu apply
• Mountable inside Next.js, Express, Hono or Fastify apps
• YAML evals for comparing models

Lobu integrations

• OpenClaw
• Slack
• Telegram
• WhatsApp
• Discord
• Google Chat
• Postgres
• GitHub
• Nix

How to get started with Lobu

1. Add Lobu to Slack with your own model keys, or open app.lobu.ai and create an organization.
2. Create an agent, add model providers and choose default models.
3. Connect chat platforms from the agent's settings and enable the skills it needs.
4. Set the worker network policy and review tool approvals before pre-approving MCP tools.
5. Invite teammates, or sync agent config from GitHub with lobu apply.

Lobu limitations

• Self-hosting means you provide ingress, DNS and the database upgrades
• Agent Protocol and A2A support are listed as not available yet
• The gateway is being rewritten, so usage tracking, audit logs and a control plane are still in progress

Lobu pricing

Lobu is open source, so you can use and self-host it for free. Paid plans start at $0 usage-based, according to the official pricing page on 24 Oct 2026.

Lobu plans

• Self-hosted

  Free

  • Open-source stack
  • Single Node process with embedded Postgres
  • You run the infrastructure

• Serverless OpenClaw

  $0.000463 /vCPU /minute

  Usage-based, billed per second while active

  • Memory $0.000231 / GB / minute
  • Storage $0.15 / GB / month
  • Idle agents cost $0
  • Sandboxed isolation per agent

Plans as listed on the official pricing page on 24 Oct 2026. Prices can change and may vary by country or exclude tax. Check the official pricing page.

Lobu alternatives

The main Lobu alternatives listed on Debuly are OpenClaw, Mem0, Zep, Letta, Supermemory and OpenBot. The table compares pricing and platforms; order is A–Z.

1. 

   Letta

   Platform, formerly MemGPT, for building stateful AI agents with persistent memory, from a terminal app, SDK or API.

   Developer platformsAutomation and agentsCoding
   FreemiumVisit Letta (official site)
2. 

   Mem0

   Memory layer for AI agents and apps, available as an open-source package or a managed platform that keeps user context across sessions.

   Developer platformsAutomation and agentsCoding
   FreemiumVisit Mem0 (official site)
3. OB

   OpenBot

   CopilotKit's open-source template for a self-hosted employee agent workspace, where each AI coworker has its own channel, browser, files and approved tools.

   Automation and agentsChat assistantsDeveloper platforms
   Open sourceVisit OpenBot (official site)
4. 

   OpenClaw

   Open-source personal AI assistant that runs on your own devices and takes actions for you through WhatsApp, Telegram, Slack, Discord and other chat apps.

   Automation and agentsChat assistantsProductivity
   Open sourceVisit OpenClaw (official site)
5. 

   Supermemory

   Memory and context API for AI agents with user profiles, a memory graph, retrieval and connectors, billed by usage.

   Developer platformsAutomation and agentsCoding
   FreemiumVisit Supermemory (official site)
6. 

   Zep

   Managed memory and context platform for AI agents, built on the Graphiti knowledge graph, with Python, TypeScript and Go SDKs.

   Developer platformsAutomation and agents
   FreemiumVisit Zep (official site)

Lobu FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: from lobu.ai · fetched 24 Oct 2026 · Vendor trademark, used only to identify the product; removed on request
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.