# Headlong on jurniti

> Headlong keeps thinking between messages — a Laude/MIT Bash microharness on its own Firecracker guest. ~$1–2/hour while the mind is up. BYOK.

Managed by jurniti; upstream project: [Laude Institute](https://github.com/laude-institute/headlong). Persistent-agency Bash microharness. The agent keeps thinking between messages. BYOK.

## What you get

- Persistent agency: messages land in a thought stream that is already running. The agent decides if and when to reply.
- Live in-browser terminal — run `headlong-init` with YOUR OpenRouter/Anthropic/OpenAI key. Set a provider spend cap first.
- Measured spend while the mind is up: about $1–2/hour (~$720–1,440/month if left 24/7). `persona stop` is the off switch.
- Persistent ~/.headlong — identity, memories, and trajectories survive restart. Public templates stay vetoed until a capture canary.
- Real Firecracker isolation. Headlong's own sandbox is off on purpose: the microVM is the boundary.

## Pricing

- **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7.
- **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in.
- **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm.

30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout.

Product page: https://www.jurniti.com/headlong