PrivateClaw – AI agents running in confidential VMs you can verify
We built PrivateClaw because the hosted OpenClaw platforms on the market today require you to trust them with plaintext. PrivateClaw removes that requirement at the hardware layer. PrivateClaw runs AI agents inside Trusted Execution Environments (TEEs), backed by AMD’s SEV-SNP standard. This means that your data is encrypted at the hardware level, enforced by the AMD Secure Processor outside the host OS trust boundary. PrivateClaw comes with inference that also runs inside TEEs, which means your prompts and completions are private as well. How it works: Each user gets a dedicated CVM…
What it does
In the maker’s words, at launch
We built PrivateClaw because the hosted OpenClaw platforms on the market today require you to trust them with plaintext. PrivateClaw removes that requirement at the hardware layer. PrivateClaw runs AI agents inside Trusted Execution Environments (TEEs), backed by AMD’s SEV-SNP standard. This means that your data is encrypted at the hardware level, enforced by the AMD Secure Processor outside the host OS trust boundary. PrivateClaw comes with inference that also runs inside TEEs, which means your prompts and completions are private as well. How it works: Each user gets a dedicated CVM (Confidential VM) — no shared tenancy. SEV-SNP provides hardware-enforced memory encryption with a per-VM key managed by the AMD Secure Processor, outside the host OS trust boundary. The hypervisor cannot read guest memory. Onboard now by running ssh privateclaw.dev in your terminal of choice. How you verify it: Our open-source CLI https://github.com/lunal-dev/privateclaw-cli is installed by default on all user CVMs and enables users to perform a 5-step verification: 1. SEV-SNP attestation — fetches a signed attestation report from the AMD PSP and validates it against AMD's root of trust 2. vTPM verification — confirms the virtual TPM's endorsement key is bound to the CVM's attestation 3. Host key binding — verifies the SSH host key you're connecting to is the one measured in the attestation report 4. Inference endpoint check — confirms the inference and inference proxy cert is bound to their respective TEE measurements 5. Access control audit — validates that only your SSH key is authorized and the cloud’s guest agent is disabled Every step is transparent and auditable, and the CLI that does this for you is open source. Today, we enable you to verify that your agent is running inside a TEE. Attestable builds are on our roadmap, which will also enable users to verify what software is running inside the TEE. Architecture: PrivateClaw runs the user CVM and inference gateway on Azure Confidential Compute, and inference itself is powered by Confidential AI's TEE-backed vLLM deployment. The launch digest for each CVM is in the attestation report, so you can verify the boot state. Binding specific userland binaries to published source is on our reproducible build roadmap. Pricing: Free tier available. Pro, with greater limits, is $69/mo. Try it: ssh privateclaw.dev
Does the same job
all alternatives →
- IPI put an AI agent on a $7/month VPS with IRC as its transport layerMar 2026 · georgelarson.me · ▲340
The stack: two agents on separate boxes. The public one (nullclaw) is a 678 KB Zig binary using ~1 MB RAM, connected to an Ergo IRC server. Visitors talk to it via a gamja web client embedded in my site. The private one (ironclaw) handles email and scheduling, reachable only over Tailscale via Google's A2A protocol. Tiered inference: Haiku 4.5 for conversation (sub-second, cheap), Sonnet 4.6 for tool use (only when needed). Hard cap at $2/day. A2A passthrough: the private-side agent borrows the gateway's own inference pipeline, so there's one API key and one billing relationship…
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
Launched alongside, April 2026
the whole month →
- AG
Thought the resources for GPU arch were lacking, so here we are
Life & fun · Apr 2026 · jaso1024.com
- IB
Built a ~9M param LLM from scratch to understand how they actually work. Vanilla transformer, 60K synthetic conversations, ~130 lines of PyTorch. Trains in 5 min on a free Colab T4. The fish thinks the meaning of life is food. Fork it and swap the personality for your own character.
AI · Apr 2026 · github.com
- BC
Life & fun · Apr 2026 · sam-burns.com
- IB
With social media and now AI, its important to keep the indie web alive. There are many people who write frequently. Blogosphere tries to highlight them by fetching the recent posts from personal blogs across many categories. There are two versions: Minimal (HN-inspired, fast, static): https://text.blogosphere.app/ Non-minimal: https://blogosphere.app/ If you don't find your blog (or your favorite ones), please add them. I will review and approve it.
AI · Apr 2026 · text.blogosphere.app