Open source infrastructure for running agentic AI on your own hardware. The senses. The memory. The tools. The interface. No mega-corp permission required.
The AI industry isn't just consolidating market share. It's consolidating the physical infrastructure underneath everything. They've bought every GPU they can find. They've cornered the training hardware. The downstream effect is already visible: RAM prices have tripled. Memory you used to buy for nothing now costs a fortune — that cost flows directly from the compute arms race these companies are running. You're already paying, whether you use their products or not.
Then they took the data. Take Anthropic. Take OpenAI. Take Google. Exceptional at operationalizing models — but the training data didn't come from nowhere. Reddit, scraped. Books pulled from authors and publishers without consent or compensation. Companies like Invisible built global networks of consultants — paid them to annotate and label, strip-mined their domain judgment, turned it into structured training data, and fed it straight to the top. The consultants were compensated. They just didn't understand they were being harvested. The scraping hasn't stopped. It won't.
Their real product isn't the model. It's you.
Here's why: they need your ongoing data to keep growing. Lock-in isn't a side effect of their business model — it's the mechanism. They have to keep you inside the bundle so the extraction can continue. Every interaction you have with their tools is training data. Every session. Every correction you make to a response that wasn't quite right. You are the product. And the product doesn't get to leave.
The way they enforce it: they wrap the LLM in a full stack of bundled services — the crawlers, the document tools, the memory, the embeddings, the shitty UIs, the threadbare editing and search, and training on your interactions — and sell the whole thing back to you as a moat. The pitch is that you couldn't possibly build this on your own. They get you hooked. Then — because you have nowhere else to go — they abuse you. The rules, the pricing, the API limits don't change every five years. They change every few months.
That ends today.
We're unbundling it. Every piece they've welded into their stack, shipped as its own tool, running on your hardware. The empires always fall. We're not building the next one.
The heavy machinery required to run agentic AI without asking permission. Open source. Self-hostable.
Run it all local // Got Docker Desktop on your machine? Every tool below runs in your own container. Cloud Run is the convenience; your hardware is the default.
Centralized AI controls the crawlers — and takes advantage of the fact that crawling the web efficiently is a massive headache for the average user. Grub is authenticatable, secure, and fully open source. Pull the code and run it on your own infrastructure, or hit the instance running on Google Cloud Run right now. If your agents need to see the web, they no longer need permission.
Running embeddings locally is possible — but downloading the models means pulling down 5 to 10 gigabytes. It turns building an AI tool into downloading a AAA game on Steam. Too much friction. SHIVVR handles chunking and embedding so your local machine doesn't have to carry the dead weight of the models.
Their cloud holds your agent's memory hostage — the moment you stop paying, the context is gone. Ferricula runs on your hardware. Memory that decays, consolidates, and dreams on its own schedule. What survives is what gets used. The rest fades. Your agent, your history.
A heavy-duty CLI orchestrator for AI workloads. Multi-provider, session-persistent, tool-rich. Spin up agents in sealed Docker containers, give them a bench of tools, and let them run. The kraken that runs the fleet below deck.
A surface where terminals, shells, and agents come together — on your hardware. Invisible AI running beside you: multiple LLMs driving your machine while you stay in control. Let them type the commands you can't be bothered to remember. Run one agent. Or let one run many.
"Thank god for Hyperia — I can't remember all these shell commands."