Box launches $0.036/hr Ubuntu VMs for AI agents

// 2h agoPRODUCT LAUNCH

Box launches $0.036/hr Ubuntu VMs for AI agents

Box provides cost-effective cloud virtual machines engineered specifically for AI agents and software factories, offering root privileges, desktop GUI, and SSH access in under two seconds. At $0.036 per hour, it serves as an economical alternative to sandboxes like E2B and Daytona for running up to 1,000 concurrent instances.

// ANALYSIS

As AI agents evolve from isolated code-interpreter sandboxes to full desktop and OS-level automation tools, infrastructure boot times and compute costs become primary scaling bottlenecks. Box directly addresses these challenges by delivering fast, low-cost virtual machines with full administrative and desktop capabilities.

–**Aggressive Pricing**: At $0.036 per hour, it dramatically reduces compute overhead for long-running or multi-agent workloads compared to existing agent sandbox tools.
–**Full OS Environment**: Provides complete root access, SSH, and desktop interaction, supporting broad agentic use cases like UI automation and complex system administration.
–**Fast & Scalable DX**: Sub-2-second VM boot times via a simple CLI command allow seamless integration into automated workflow pipelines.

// TAGS

boxdevtoolartificial-intelligenceagentcloud-computinginfrastructure

DISCOVERED

2h ago

2026-07-22

PUBLISHED

7h ago

2026-07-22

RELEVANCE

8/ 10

AUTHOR

[REDACTED]

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

UPDATE28m ago

MōBrowser integrates AI agents into dev loop

MōBrowser bridges AI agents with the application runtime environment, enabling agents to read framework documentation, build and launch applications, inspect interfaces, interact with running apps, diagnose problems, and verify code changes automatically. By embedding real-time execution feedback loops into the development workflow, it advances autonomous software engineering beyond isolated code generation.

RESEARCH29m ago

USTC debuts PUMA to curb LLM overthinking

PUMA is a training-free diagnostic framework developed by USTC researchers to detect and mitigate cognitive stagnation in Large Reasoning Models. By analyzing geometric momentum and entropic uncertainty in latent space, PUMA enables inference engines to adaptively truncate redundant reasoning trajectories and lower token consumption.

MODEL52m ago

Google solicits feedback on Gemini 3.6 Flash

Jack Woth posted an inquiry on X requesting honest developer feedback on the performance and cost-efficiency of Gemini 3.6 Flash and Gemini 3.5 Flash-Lite. The goal of soliciting this community input is to ensure that the newly updated models reliably and efficiently handle real-world developer workflows and practical applications.