E2B sandbox

Per second, wall-clock, for provisioned vCPU and RAM. Storage included.

Cost / code-agent task

$0.0138

2 vCPU / 4 GiB sandbox for 5 minutes at 10% CPU utilization (most time spent waiting on the LLM). Mirrors Vercel's own "AI code validation" example.

Free tier

Hobby: $0/mo, one-time $100 usage credit, no credit card.

source ↗ · 2026-10-17

Isolation

Firecracker microVM

source ↗

Rates

Item	Rate	Source	Fetched
vCPU	$0.000014 /vCPU-second	e2b.dev ↗	2026-10-17
RAM	$0.0000045 /GiB-second	e2b.dev ↗	2026-10-17
Storage	included	e2b.dev ↗	2026-10-17

Plans

Plan	Monthly fee	Included	Concurrency	Max session	Source
Hobby	$0	One-time $100 credit	20	1 hour	source ↗
Pro	$150	Usage billed separately	100 (add-ons to 600 / 1,100)	24 hours	source ↗
Enterprise	Custom	$3,000 monthly minimum	Custom	Custom	source ↗

Limits & details

vCPU / RAM per sandbox: 1–8 vCPU, 1–8 GiB (default 2 vCPU / 4 GiB) ↗
GPU: None (CPU-only) ↗
SDKs: Python, JS/TS ↗
Pricing page: https://e2b.dev/pricing
Terms: https://e2b.dev/terms
Last verified: 2026-10-17

Cold start

No Coldtally benchmark runs published yet. For broad speed coverage see ComputeSDK's open benchmarks (MIT-licensed data). Our methodology explains how we use them.

Changes