your ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/week
Live now· #4 this week
jilebi

jilebi

developer machines in the cloud for AI

Visit website

Subscription87 impressions#4 this week1 comment+2 today

Comments

>log in to comment
  • Datrova Software[maker]· 17h ago

    Hello DevHunt community! We built Jilebi to give developers cheap, performant Linux machines in the cloud for AI workloads. You can spin one up in seconds, it auto‑sleeps when idle and you only pay for runtime, with secure isolation and automatic routing. Would love your thoughts 😊

Isolated Linux Machines pre-installed with any software you need to ship. Pay only for the time they run. the pricing does not include stopped machines that cost $0.0001/GB·hr You can spin up a machine in seconds and pay only for the time it runs. While it sleeps, only pay for disk. It also performs well - with dedicated CPU, memory, and disk for each machine. After configuring your machine, clone it to spin up more machines with the same setup in seconds. Launch a machine in seconds, pre-installed with the tools you choose. No waiting on infrastructure.
  • Launch a machine in seconds
  • Auto‑sleep when idle, wake in milliseconds
  • Secure isolated sandbox per machine
  • Automatic routing by username, machine and app name
  • Works with standard Docker, SSH, no special CLI

Jilebi provides isolated Linux machines in the cloud for AI workloads, billed only for runtime and storage.

for
Developers and teams needing on-demand AI compute environments
pricing
paid
works withLinuxDockerSSHSCPcurl

Key features

6 features of jilebi
  • Instant spin-up — Launch a pre-installed Linux machine in seconds without waiting for infrastructure.
  • Auto-sleep — Machines automatically sleep when idle, charging only for active runtime.
  • Dedicated resources — Each instance gets its own CPU, memory, and disk for predictable performance.
  • Route-based access — Access apps via username/machine/app subdomains instead of IPs or ports.
  • Standard tooling — Use Docker, SSH, SCP, and curl just like on a normal machine—no special CLI needed.
  • Cloneable configs — Clone a configured machine to replicate setups instantly.

Use cases

  • Run AI agents that need isolated, reproducible environments
  • Spin up temporary compute for model training and shut down when finished
  • Share a development sandbox with teammates via secure routes
  • Test different AI stacks (e.g., Claude Code, Hermes) without local installation

jilebi FAQ

How is usage billed?+

You pay per vCPU hour, GB of memory hour, and GB of disk hour; stopped machines only incur disk storage cost.

Is there a free tier?+

No free tier is mentioned; pricing is usage-based.

Can I clone a machine configuration?+

Yes, after configuring a machine you can clone it to create identical instances quickly.

How do I access a running machine?+

SSH directly to a routed subdomain using your username, machine name, and app name.

Summarized by DevHunt from jilebi.ai · Nov 2, 2026. Details may change; check the official site.

More dev tools

  1. OffiAIOffiAI · A free Mac app that gives every AI coding agent a desk in a 3D office