E2B
E2B is an open-source infrastructure service that provisions isolated Linux sandboxes in the cloud for AI agents and code-execution workflows. Each sandbox is a lightweight microVM that can be spawned, used to run commands, paused, resumed, and shut down via SDK calls. Customers include teams building coding agents, data-analysis tools, RL environments, and computer-use agents.
| Pricing model | Pro plan at $150/month plus per-second vCPU/RAM usage; Hobby tier free with a one-time $100 credit |
|---|---|
| Pricing The headline compute price as the provider states it. Sorting ranks prices stated per vCPU over time, converted to hourly. Other pricing models sort after, unranked. | $0.000014/vCPU-second (≈$0.0504/vCPU-hour); RAM $0.0000045/GiB/second (≈$0.0162/GiB-hour) |
| Free while idle Whether a stopped, paused, or sleeping sandbox costs nothing. | ✓ |
| Elastic Resources flex with what the sandbox actually uses, and the bill follows: yes when the provider bills on active or observed use, no when capacity is reserved or allocated and billed while it runs. | ✕ |
| Memory snapshots Whether RAM state survives a pause and resume, not just disk. | ✓ |
| Wake on request Whether a stopped sandbox wakes automatically on inbound traffic. | ✓ |
| Start / resume Typical time from create or resume to running. Sorting ranks stated times, converted to seconds; claims with no figure sort after, unranked. | ~1 s to resume a paused sandbox; create time not stated |
| Max runtime The longest a sandbox may run. | 24 h on Pro plan; 1 h on Hobby (free) plan |
| Isolation The isolation technology between sandboxes. | Firecracker microVM |
| GPUs Whether GPU instances are available for sandboxes. | ✕ |
| Docker Whether Docker containers can run inside the sandbox. | ✓ |
| SDKs | Python, JavaScript, TypeScript |
✓ yes · ✕ no · – no cited public fact.
Docker support requires at least 2 vCPUs and 2 GiB RAM per the official Docker template. Firecracker lacks PCIe support, so GPU passthrough is not available. Pausing a sandbox via the SDK persists both filesystem and in-memory state (processes, variables), so the next 'connect' resumes from the same state.
Sources
- https://e2b.dev/docs/template/examples/docker
- https://github.com/e2b-dev/e2b
- https://github.com/e2b-dev/e2b#sdks
- https://thesequence.substack.com/p/the-sequence-ai-of-the-week-698-how
- https://e2b.dev/
- https://e2b.dev/docs/sandbox/auto-resume
- https://e2b.dev/docs/sandbox/filesystem-only-snapshots
- https://e2b.dev/docs/sandbox/persistence
- https://e2b.dev/docs
- https://e2b.dev/docs/quickstart
- https://pub.towardsai.net/e2b-ai-sandboxes-features-applications-real-world-impact-75e949ded8a7
- https://e2b.dev/pricing
- https://www.beam.cloud/blog/e2b-pricing-explained
- https://www.morphllm.com/e2b-pricing
- https://checkthat.ai/brands/e2b
- https://infragap.com/tools/e2b/
- https://infragap.com/compare/e2b-vs-daytona/
- https://e2b.dev/blog/docker-e2b-partner-to-introduce-mcp-support-in-e2b-sandbox
- https://e2b.dev/blog/firecracker-vs-qemu
- https://e2b.dev/docs/template/examples/docker.md
- https://e2b.mintlify.app/docs/sandbox/auto-resume.md
- https://docs.e2b.dev/faq/gpu-support
- https://docs.e2b.dev/template/examples/docker
Last verified 2026-09-26. Corrections land through the update log.