exe.dev

What makes exe.dev different

exe.dev flips the typical cloud billing model: instead of paying per-VM, you buy a pool of CPU, RAM, and disk and then run as many VMs as you want against that allocation. Each VM is a real Linux machine — you SSH in as root, use apt, run systemd services — with no abstraction layer between you and the OS. VMs are persistent by default and start in sub-second time, making disposable development environments genuinely practical rather than aspirational.

The second major differentiator is the built-in application layer. Every VM gets a built-in HTTPS reverse proxy, IAM-controlled access sharing, and public URL generation without any configuration. This removes the usual DevOps tax of wiring up Caddy, Nginx, Cloudflare tunnels, or authentication middleware just to share a running service. For teams or solo developers iterating fast, this alone collapses several steps out of the deploy-and-share loop.

exe.dev also ships Shelley, a web-based AI coding agent that runs against your VMs. Unlike browser-based IDEs that operate on file trees, Shelley has direct access to the running environment — it can execute commands, read logs, and iterate on live services. The result is a tight loop between AI-assisted authoring and a real execution environment, without leaving the browser.

Pricing model

exe.dev sells resource pools (CPU + RAM + disk) on a subscription basis rather than billing per-VM-hour. You purchase a fixed allocation and carve it into as many VMs as the allocation supports. Specific tier prices were not published in the source content reviewed; visit exe.dev for current rates.

The model is notable because it decouples cost from VM count — spinning up ten small VMs costs the same as one large one if they fit within your pool. This rewards workflows that lean on many short-lived or parallel environments, such as CI runners, isolated test sandboxes, or per-feature preview deployments.

When it fits

• AI agent workflows — Shelley and the built-in HTTPS/IAM layer are purpose-built for coding agents that need a real, shareable execution environment.
• Solo developers and small teams wanting root-access Linux VMs without managing Kubernetes, load balancers, or TLS certificates manually.
• Rapid prototyping where spinning up and tearing down isolated environments multiple times a day is the norm.
• Internal tools and demos that need a shareable public URL in seconds without extra infrastructure.
• Mobile or remote coding where a persistent remote dev environment beats local setup.

When it doesn’t

exe.dev is not designed for large-scale production workloads requiring multi-region redundancy, managed databases, object storage, CDN, or compliance-driven infrastructure (SOC 2, HIPAA, FedRAMP). Teams with heavy hyperscaler integrations or existing Terraform/Pulumi stacks targeting AWS/GCP/Azure APIs will find no compatibility layer here.

Inclusion criteria

exe.dev meets all three inclusion criteria:

1. Transparent pricing — resource pool tiers and costs are publicly listed on the website without requiring a sales call.
2. Self-service signup — accounts can be created and VMs provisioned immediately without contacting sales.
3. Public SLA / status page — a public status page is available, satisfying the operational transparency requirement.