Dedalus Labs
Full Linux machines for AI agents in under 250ms
San Francisco, United States · Founded 2025
- Headquarters
- San Francisco, California
- Employees
- 1–10
- Founded
- 2025
- Business Model
- B2B
- Website
- dedaluslabs.ai
About
Dedalus Labs is a Y Combinator-backed (S25) infrastructure company building the compute substrate for AI agents. Founded in 2025 by Catherine Di and Windsor Nguyen, and headquartered in San Francisco, the company provides Dedalus Machines—persistent, full Linux virtual machines that spin up in under 250 milliseconds, enabling developers to run long-lived, stateful AI agents with zero cold starts. Unlike traditional sandboxes and container runtimes, Dedalus Machines offer VM-level kernel isolation, root access, support for any package manager (apt, pip, npm, cargo), GPU/CUDA support, and a persistent filesystem that survives across sessions. Compute is billed only during active usage, making it significantly cheaper than competing ephemeral sandbox solutions. The platform targets developers building agentic AI applications including code generation, ML training, web scraping, and CI pipelines, and also provides an MCP gateway for connecting agents across any model and provider.
Summary
Dedalus Labs is an Artificial Intelligence company based in San Francisco, United States, founded in 2025.
Similar Companies
Cursor is the best way to code with AI, and built to make you extraordinarily productive.

Mistral is on a mission to push AI forward, to serve the open community and their enterprise cust...
Provides AI inference hardware and cloud platform for enterprises, using RDUs and SambaCloud to r...

UiPath is an AI-enhanced end-to-end automation platform.
ML inference infrastructure for deploying and serving AI models at scale, performantly and cost-e...

AI hardware company building wafer-scale processors for deep learning training and inference