Self-Hosted AI Assistants
This section covers three self-hosted AI assistant projects: OpenClaw (gateway-first, messaging integrations), NemoClaw (NVIDIA's enterprise security layer for OpenClaw), and Hermes Agent (agent-first, persistent cross-session memory plus a learning loop that builds reusable skill documents). All three run on Liquid Web infrastructure.
Two events in early 2026 changed the OpenClaw security landscape. OpenClaw's creator joined OpenAI in February 2026, raising enterprise questions about data sovereignty and long-term independence. Security researchers simultaneously identified 341 malicious skills in the ClawHub marketplace and over 135,000 exposed OpenClaw instances. NemoClaw (announced at GTC 2026 on March 16) is NVIDIA's direct response: process-level sandboxing, network egress policies, and audit logging that contain any compromised skill or misconfigured agent. If you are running OpenClaw for any sensitive workload, read What Is NemoClaw before installing.
Which assistant fits your use case?
| Use Case | Recommended | Why |
|---|---|---|
| Personal productivity + messaging integrations | OpenClaw | Connects to WhatsApp, Telegram, Slack, Discord, and 16 more platforms |
| Repeated complex workflows | Hermes Agent | Persistent memory + learning loop means it gets faster over time |
| Healthcare or enterprise compliance | NemoClaw + OpenClaw | Process-level sandboxing, audit logs, policy enforcement |
| Local GPU inference (70B models) | Hermes + Ollama or OpenClaw + Ollama | Both support Ollama; Hermes's learning loop compounds on local inference |
| Lowest resource footprint | OpenClaw | ~180 MB idle vs Hermes's ~2 GB idle |
Not sure? Read the full comparison.
OpenClaw
OpenClaw (github.com/openclaw/openclaw) is a TypeScript AI assistant with ~359k GitHub stars and an MIT license. It runs on your own machine or server, connects to 20+ messaging platforms, and supports local GPU inference and cloud API backends.
- What Is OpenClaw?: features, integrations, system requirements
- Self-Host OpenClaw on Liquid Web VPS: step-by-step Docker setup
- Docker Compose Stack: Redis, Caddy, backups, measured RAM footprint
- GPU Inference on Liquid Web: tier selection and Ollama/vLLM config
- HIPAA-Compliant Deployment: PHI guardrails and Liquid Web HIPAA setup
- Skill Development Guide: ClawHub SKILL.md format, test harness, publishing
- Monitoring & Observability: Prometheus, Grafana, alert runbook
NemoClaw
NemoClaw (github.com/NVIDIA/NemoClaw) is NVIDIA's security layer for OpenClaw. Announced at GTC 2026, it runs OpenClaw inside an OpenShell sandbox with process-level isolation, network egress policies, and audit logging. Currently in Alpha.
- What Is NemoClaw?: architecture, inference modes, alpha caveats
- NemoClaw Policy Cookbook: Colang policy recipes for DLP, HIPAA, and skill gating
Hermes Agent
Hermes Agent (github.com/nousresearch/hermes-agent) is an open-source AI assistant from Nous Research (MIT, ~128k stars, v0.12.0). Unlike OpenClaw's gateway-first architecture, Hermes is agent-first: it plans, executes, reflects, and retries. It stores episodic and semantic memory in PostgreSQL so it remembers your workflows across sessions. After completing a novel task with confidence ≥ 0.7, it writes a reusable skill document, so it gets faster the second time.
- What Is Hermes Agent?: architecture, memory types, system requirements
- Docker Compose Stack: PostgreSQL, pgvector, Caddy, restic backup
- Learning Loop & Skills: guide and measure the learning loop; inspect skill documents
OpenClaw vs Hermes Agent vs alternatives
See vs Alternatives for a full comparison across OpenClaw, NemoClaw, Hermes Agent, Open WebUI, AutoGen, CrewAI, Dify, and more.
Resource footprint at a glance
| Tool | Idle RAM | Recommended Liquid Web tier |
|---|---|---|
| OpenClaw (remote LLM API) | ~180 MB | 4 GB Cloud VPS |
| OpenClaw + Ollama 7B | ~10 GB | L4 GPU server (24 GB VRAM) |
| Hermes Agent (remote LLM API) | ~2 GB | 8 GB Managed VPS |
| Hermes Agent + Ollama 70B | ~52 GB+ | L40S GPU server (48 GB VRAM) |
Prices subject to change, so verify at liquidweb.com/vps-hosting/managed-vps/ and liquidweb.com/gpu-hosting/.
OpenClaw is an open-source personal AI assistant (MIT license, TypeScript) with ~359k GitHub stars as of April 2026. It runs on your own hardware or VPS, supports 20+ messaging platforms, local GPU inference via Ollama, and 5,700+ community skills through ClawHub.
NemoClaw is NVIDIA's open-source plugin (Apache 2.0) that runs OpenClaw in a sandboxed OpenShell environment with enterprise security controls. It was announced at GTC 2026 and is currently in Alpha.
Hermes Agent is an open-source AI assistant from Nous Research (MIT, ~128k stars, v0.12.0). Unlike OpenClaw, it maintains persistent memory across sessions and runs a learning loop that generates reusable skill documents from successful task completions.
No. Both run on a VPS using a cloud API backend like Claude or OpenAI. GPU is only needed for fully local inference via Ollama or vLLM, which eliminates cloud API costs and keeps all data on your hardware.
Not by default. For HIPAA compliance you need NemoClaw's sandboxing, a Business Associate Agreement (BAA) with your hosting provider, encrypted storage, and audit logging. Liquid Web offers HIPAA-compliant dedicated servers and signs a BAA; pricing is quote-based.
