Dedalus Labs

What is Dedalus Labs

Dedalus Labs provides Dedalus Machines, which are described as the fastest persistent sandboxes for AI agents. These machines allow users to spin up long-running, stateful agents that can operate continuously, with boot times under 50 milliseconds.

How to use Dedalus Labs

1. Install the Dedalus CLI: For macOS users, the command is brew install dedalus-labs/tap/dedalus.
2. Set your API key: Export your API key using the command export DEDALUS_API_KEY=<YOUR_API_KEY>.
3. Create a machine: Use the command dedalus machines create --vcpu 1 --memory-mib 1024 --storage-gib 10 to create a machine.
4. Run a command: Execute commands on your machine using dedalus run <machine-id> -- <command>.

Features of Dedalus Labs

• Fast Boot Times: Machines boot in under 50ms.
• Persistent Sandboxes: Long-running, stateful agents that never sleep.
• VM-Level Isolation: Provides kernel-level and hardware-level isolation for security.
• Full Linux Environment: Offers unrestricted Linux environments with systemd, root access, and package managers.
• Persistent Storage: Memory, volumes, and compute survive restarts.
• Scale-to-Zero: Pay only for active compute time; sleep is free.
• Per-Second Billing: Billed per-second for active CPU time.
• GPU Support: Available for machine learning workloads.
• Nested Virtualization: Allows for building custom sandboxes.
• Live VM Migration: Enables zero-downtime migrations.

Use Cases of Dedalus Labs

• AI Agent Hosting: Running long-running, stateful AI agents.
• Development Environments: Providing fast, isolated development environments.
• Code Execution: Securely executing code in a sandboxed environment.
• Machine Learning Workloads: Utilizing GPU support for ML tasks.

Pricing

• Free Tier: Offers 50 free hours of compute per month, with no credit card required.
• Pro Tier: Starts at $20/month, offering higher rate limits and priority support. Billing is per-second for active compute time. Storage is priced separately.

FAQ

• What is Dedalus Labs? Dedalus Labs builds Dedalus Machines, which are VMs that boot in 50ms with full Linux, persistent storage, and VM-level isolation. The platform provides production compute infrastructure for AI agents with per-second billing and scale-to-zero pricing.
• What are Dedalus Machines? A Dedalus Machine is a full Linux VM that boots in 50ms. Each machine has its own guest kernel with hardware-enforced VM-level isolation, persistent storage, and per-second billing. Compute and storage are disaggregated.
• How fast do Dedalus Machines boot? Dedalus Machines boot in 50ms using snapshot restore, significantly faster than traditional methods.
• What isolation model does Dedalus use? Dedalus uses VM-level isolation with hardware enforcement via privilege ring transitions and Extended Page Tables.
• How does Dedalus Machines pricing work? Users receive 50 free hours monthly. The Pro tier starts at $20/month, with per-second billing for active CPU time. Storage is persistent and billed separately.
• What can I run on a Dedalus Machine? Anything that runs on Linux, including systemd, any package manager, runtime, or binary. GPU support is available.