FlashVMM launches local MicroVM sandbox for AI agents
Start here
FlashVMM has released a local sandbox that gives each AI agent task its own FlashVM, a custom virtual machine that the team says starts warm in milliseconds and is discarded when the task ends. It installs via a shell or PowerShell script on macOS, Windows and Linux, requires no account or cloud, and runs existing agents such as Claude unchanged inside a VM that sees only the project folder. The product also offers an MCP server with 21 tools for creating, forking, snapshotting and tearing down VMs. A GPU pool loads a model once on the host GPU and serves requests to allowed VMs, using Metal on Apple silicon and Vulkan on Linux and Windows. The millisecond warm-start claim comes from the team and has not been independently verified.
Generated from reports · updated 17 h ago
Report timeline
Follow the reports to see the event from different sides.
- Hacker News · AIFlashVMM launches a local MicroVM sandbox that isolates AI agent tasks
FlashVMM is a local sandbox that gives each AI agent task its own FlashVM, a custom virtual machine the team says starts warm in milliseconds and is discarded when the task ends. It installs via a shell or PowerShell script on macOS, Windows and Linux, runs existing agents like Claude unchanged inside a VM that sees only the project folder, and offers an MCP server with 21 tools for creating, forking, snapshotting and tearing down VMs. A GPU pool loads a model once on the host GPU and serves requests to allowed VMs, using Metal on Apple silicon and Vulkan on Linux and Windows.
Heat for this event
- Current, comparable range
- 6
- Peak, comparable range
- 10November 16 00:00
- Change in 24 hours
- –
The trend compares the same participants over hours observed continuously, so its range can be narrower than the current heat count. Hover or tap the chart for hourly heat; the arrow keys move through it.