You’d think I’d be the last person to say “sandboxes are dead.” But I’ll be the first to tell you that sandboxes aren’t the whole story for agentic infrastructure. A lot of the loudest takes here come from people who haven’t built agents that operate at scale. The reality is less convenient: there’s no single tool that fits every task. What matters is knowing what to use, when to use it, and which tradeoffs you’re making. Everyone feels like they need a strong take. It has to be A or B. In the real world at real scale, it’s usually some combination of both. The real world is messy. Infrastructure at scale is full of constraints you don’t see in a demo. If you haven’t dealt with them, it’s easy to mistake what worked for your agent for what should work for everyone. At @e2b, we believe agents will increasingly make these infrastructure decisions themselves. They’ll choose the resources and environments they need for the task, with the underlying infrastructure handling orchestration, isolation, and control. The only way to survive in AI is to keep building for where things are going. By the time you’ve built for what agents can do today, they’re already capable of more. And the easiest and fastest way to be wrong is to think your solution is the one. Sandboxes are part of that future. There’s a lot more to build around and alongside them. That’s what we’re building at @e2b

Give @Meta's Muse Code an @e2b sandbox and its native video understanding yields an efficient parallel video-processing pipeline. We handed the same demo recording to Muse, Claude Code, and Codex, each in its own E2B sandbox, and asked for a step-by-step breakdown. Muse Spark 1.3 takes the video natively. The others have to shell out to ffmpeg, extract frames, and re-send every image on every turn. In this clip, Muse finished up to 9.3× faster at up to 99.4% lower cost. Spin up one isolated sandbox per clip, and Muse applies edits across hundreds of videos in parallel. Try it: e2b sbx create muse Docs: docs.e2b.dev/agents/muse

Replicas cut latency for customers running heavy workloads by more than 60% by changing one thing: the computer underneath every coding agent. Before E2B, the team tried building sandbox infrastructure themselves on raw EC2, then with five other sandbox providers. Of every option Replicas tested, E2B was the fastest and most reliable to build on. Today, @tryreplicas keeps warm pools of E2B MicroVM sandboxes preloaded with each customer’s repo and development environment to achieve the best possible user experience for their customers. "I originally started with a competitor, but the search for something stable kept dragging on right as the team was heading into the end of its YC batch and preparing to fundraise. The sandbox was basically the last thing I wanted to think about.” - Connor Loi, Founder, Replicas Read the full case study: e2b.dev/blog/replicas

Vercel's AI SDK 7 ships HarnessAgent, an API for running established agent harnesses like Claude Code, Codex, and Pi. Write the agent once, then use whatever harness is best for your use case. Try building a creative production agent with @pidotdev using the @vercel AI SDK: one image model makes the hero shot, then an E2B sandbox runs Python to cut the backdrop, render every specified size, overlay copy, audit WCAG contrast and palette share, and serve the finished ad set on a live URL. Full example in the cookbook: github.com/e2b-dev/e2b-cookb…

Two new integrations with @e2b: @herdrdev & @colemurray's background agents. Use them with E2B to give your agents persistent, fully customizable computers built for stateful code execution. Here's what you can build with those integrations 👇

How do you hand over sensitive insurance data to agents without losing control of it? That's the question Effective AI had to solve. Their agents build rating plans, compare competitor filings, and run portfolio analyses. Every workload has to respect strict data governance: SOC 2 Type II controls, tenant isolation, configurable retention, and no training on customer data. E2B extends those same guarantees down into code execution: every agent workload runs in its own MicroVM-isolated sandbox, provisioned only with the resources it's authorized to touch, so sensitive data stays protected even as work fans out across multiple agents in one workflow. Today, Effective helps insurance organizations collectively writing more than $180 billion in direct written premiums launch and maintain insurance products using E2B. Read the full case study: e2b.dev/blog/effectiveai.

Going from prompt to an app you can serve to users is magical. It's also very hard to pull off at a scale of 1M+ users. Anything's Marcus Lowe and his team has been able to build something incredible in a very short amount of time. E2B has been a part of that: absorbing the concurrency spikes that come with virality, keeping apps warm so builders can resume editing the second they log back in, and serving the finished apps once they ship them. "I still don't know how you guys do that" - my favorite quote from this case study. Shout-out to both team @anything and @e2b for building something out of this world.