BonrockWho's Actually Making Money with OpenClaw
OpenClaw started as a hobbyist project. Now people are using it to automate businesses, build products, and offload work they used to pay for. Here's what that actually looks like.
The people getting the most value from OpenClaw aren't the ones posting about it.
They're the freelancers whose dev loops now involve less context-switching. The consultants whose intake process runs without their attention. The developers who deployed an iOS app to TestFlight while watching Netflix. The people who looked at the tool, figured out what their specific friction was, and built something that removed it.
This is what real adoption looks like: not product launches, but quiet compounding. Here's what's actually being built.
Most of the Value Is in Replacing Human Time on Boring Work
The most common pattern: taking repetitive work that required touching external systems and building an agent pipeline that handles it instead.
A tax consultant in the community built intake automation. PDFs arrive by email. The agent collects them, preps the documents, and handles the monthly accounting intake process. The consultant handles only the judgment work. The agent's value is in removing the overhead that was limiting how many clients the consultant could serve.
@marchattonhere automated Tesco grocery shopping. The agent takes a weekly meal plan, generates a shopping list, books a delivery slot, and confirms the order through browser automation. No integration contract with Tesco required; the agent interacts with the website the way a person would, which means it can do anything a person can do with a browser.
@George5562 automated school meal booking through ParentPay, a UK school payments system that's notoriously unpleasant to use manually. The implementation uses mouse coordinates for clicking table cells: direct, specific, unglamorous, and saves real time every week.
@attol8 built a job search skill in 30 minutes using the JSearch API. It searches listings, matches against CV keywords, and returns relevant opportunities. It's the kind of tedious work people either do manually or pay someone else to handle.
What connects all of these: the value is in replacing human time on tasks that have a clear pattern but require navigating real systems. An agent connected to a messaging app that can make decisions and click things turns out to be useful for a lot of work that never had a good software solution.
How Developers Changed Their Relationship With Their Own Infrastructure
Developers have found uses that are less about automation and more about changing the interface to their work.
@banknokia set up a pipeline where a coding agent finishes a change, opens a PR, and OpenClaw then reviews the diff and replies in Telegram with suggestions and a merge verdict. The developer stays in Telegram; code reviews happen asynchronously, without opening the repo.
@coard built a complete iOS app with maps and voice recording features, deployed it to TestFlight, through Telegram chat. Never opened his laptop. The agent ran Xcode builds, handled code signing and provisioning, and pushed the release. The constraint of needing to sit at a keyboard to get development work done stopped applying.
@davekiss rebuilt his entire personal website while watching Netflix. Migrated 18 posts from Notion to Astro, moved DNS to Cloudflare. The agent handled the execution over Telegram while he was doing something else.
@adam91holt is running 14+ agents under one gateway, with a Claude orchestrator delegating to Codex workers. He documented the setup publicly: role separation, model selection, sandboxing, webhook configuration, heartbeat setup, delegation patterns. It's a production multi-agent system.
Some People Skipped Personal Productivity and Built Products
Some people built products with OpenClaw as infrastructure.
Clawdia Phone Bridge (@alejandroOPI) connects Vapi voice calls to an OpenClaw agent via HTTP bridge. You can call your agent. It responds in near-real-time with voice output, backed by an agent that has full system access.
SNAG (@am-will) is a screenshot-to-Markdown tool. Hotkey a region of your screen, Gemini vision processes it, and you get Markdown in your clipboard. Useful for developers pulling content from non-copyable sources.
Karakeep Semantic Search (@jamesbrooksco) adds vector search to Karakeep bookmarks using Qdrant and embeddings. It's a standalone search tool for personal knowledge management; works independently, uses the OpenClaw infrastructure pattern.
StarSwap is a marketplace for astronomy equipment (buying and selling gear) built around OpenClaw. A full vertical commerce product.
Agents UI (@kitze) is a desktop app to manage skills and commands across Claude Code, Codex, and OpenClaw, built specifically to serve the OpenClaw user base.
These aren't "things you do with OpenClaw." They're products where OpenClaw is part of the stack, and the people behind them are charging for, or preparing to charge for, the end product.
Home Automation Keeps Appearing
It comes up consistently enough to mention separately.
@joshp123 built GoHome, a full home automation system with OpenClaw as the conversational interface and Grafana dashboards for monitoring. His Roborock vacuum is part of the same system. @antonplex has Claude Code discovering device control interfaces and OpenClaw managing ongoing air quality monitoring. @ngutman packaged the whole setup as a Home Assistant add-on with SSH tunnel support for remote access.
The pitch is specific: natural language control of your home, with an agent that remembers context across conversations, can chain multi-step actions, and is reachable through the same messaging app you use for everything else. "Tell the agent to run the vacuum at 10pm when the house is empty" beats navigating smart home apps.
What "Making Money" Actually Means Here
Direct monetization is a minority. Most OpenClaw users aren't selling access or building revenue-generating products on top of it.
What they're doing is arbitraging their own time against API and VPS costs. A $20/month VPS plus maybe $50-100/month in API costs, against the value of work the agent handles. The people running intake automation, automated research pipelines, deployment watchers: they're getting hours back.
For some of them, those hours have a clear dollar value: client billing, development throughput, the ability to service more customers. For others it's less direct but no less real.
The people who built actual products are a smaller group, but they exist. The ecosystem is early enough that the category hasn't been defined yet, which is exactly when it's worth paying attention to.
The people who figured this out first are already months into compounding it.
Found this helpful? Pass it on.