# E2B

Firecracker microVMs built for agent code. Pause keeps memory and disk until you resume.

Route to it: `--provider e2b`.

Compare it with others: [/providers/compare?p=e2b](/providers/compare?p=e2b).

## Price

| Size | Asked for | Per hour | Per month | Runs as | Plan |
| --- | --- | --- | --- | --- | --- |
| S | 1 vCPU · 2 GiB | $0.099/hr | $72/mo | 1 vCPU · 3 GiB · 10 GiB | $0 · $100 credit |
| M | 2 vCPU · 4 GiB | $0.198/hr | $145/mo | 2 vCPU · 6 GiB · 10 GiB | $0 · $100 credit |
| L | 4 vCPU · 8 GiB | $0.396/hr | $289/mo | 4 vCPU · 12 GiB · 20 GiB | Pro $150/mo + raise |
| XL | 8 vCPU · 16 GiB | Not offered | Not offered | — | — |

A month is 730 hours of running. E2B bills CPU the whole time a sandbox runs, so it is always 100%.

## Limits and features

What works through Sandbox Router:

| Feature | Value | Details |
| --- | --- | --- |
| Isolation | microVM | Firecracker |
| Snapshots | Memory | and fork |
| Kept after stop | Disk + memory |  |
| When idle | — |  |
| Max session | No limit | VM renewed hourly |
| Max vCPU | 4 | Pro $150/mo + raise, estimated |
| Max memory | 12 GiB | Pro $150/mo + raise, estimated |
| Max disk | 20 GiB | Pro $150/mo + raise, estimated |
| Sizing | XS–2XL, or custom |  |
| Custom image | Yes |  |
| GPU | No |  |
| Self-hosting | Yes | open source, BYOC |

- † On your own key, the provider's limits for your plan apply. E2B's Hobby plan fits up to M: L needs E2B Pro and a limit raise. XL (8 vCPU · 16 GiB) is over every provider's limits through Sandbox Router.
- Shape not confirmed: Sandbox Router may boot E2B's base template, 2 vCPU · 512 MiB · 10 GiB at every size, so its prices and limits here are estimates.

Pricing sources, checked 6 Nov 2026: [e2b.dev/pricing](https://e2b.dev/pricing), [Price calculator FAQ](https://docs.e2b.dev/faq/calculate-sandbox-price), [Billing docs](https://docs.e2b.dev/billing).

## Benchmarks

From [ComputeSDK](https://www.computesdk.com/benchmarks/sandboxes), which runs the same test against every provider. Latest run Nov 6, 2026.

- Burst TTI score: 87.3. Rank 16 of 30 on the latest run, up 1.1 from Oct 30. ([source](https://www.computesdk.com/benchmarks/sandboxes/burst-tti/e2b))
- Build time: 99.7 s. Rank 23 of 35 on the latest run, 23.9 s slower than Oct 30. Failed Oct 16 to 23. ([source](https://www.computesdk.com/benchmarks/sandboxes/dax/e2b))

| Run | Burst TTI score | Build time |
| --- | --- | --- |
| Nov 6, 2026 | 87.3 | 99.7 s |
| Oct 30, 2026 | 86.2 | 75.9 s |
| Oct 23, 2026 | 84.6 | Failed |
| Oct 20, 2026 | 85.8 | Failed |
| Oct 16, 2026 | 86.3 | Failed |
| Oct 9, 2026 | 85.3 | 72.7 s |
| Oct 2, 2026 | 85.9 | 94.3 s |
| Sep 25, 2026 | 86.2 | 77.1 s |