[ENGINEERING]
Choose the right storage for your Blaxel agents
Blaxel agents have three storage options: the sandbox filesystem, volumes, and Agent Drive. Understand what each option does and when to use it.
July 06, 2026 • 12 minutes reading.
[ENGINEERING]
How to slash sandbox memory usage by 75% using OverlayFS
We improved sandbox performance with OverlayFS and EROFS technologies, cutting memory usage by 75%. Discover how we did it.
May 14, 2026 • 7 minutes reading.
[ENGINEERING]
Stop paying for idle: how we built per-second metering for Blaxel Sandboxes
How we built event-driven usage metering for Blaxel Sandboxes, AI Hosting and Batch Jobs, to ensure credit-based billing only for when compute is really used.
May 14, 2026 • 10 minutes reading.
[ENGINEERING]
How we optimized MCP servers for serverless (or: From SSE to WebSockets)
A deep-dive into our journey making MCP servers compatible with WebSockets.
May 14, 2026 • 12 minutes reading.
[ENGINEERING]
We slashed our network latency to under 50 milliseconds
Our next-gen infra reduces request latency to sub-50 ms, enabling near-instant agentic responses across the network.
May 14, 2026 • 10 minutes reading.
[ENGINEERING]
What if Lambda and EC2 had a baby?
Building for agentics requires more than just containers. We moved to bare metal to give agents instant-launching persistent sandboxes. An anatomy of our runtime.
May 14, 2026 • 12 minute read.
[ENGINEERING]
Understand the lifecycle of a Blaxel sandbox
Learn how sandboxes are built, managed, billed, and cleaned up behind the scenes.
May 13, 2026 • 12 minutes reading.