# Tensorlake vs Vercel Sandbox: prices, limits and features.

**Their prices for 2 vCPU and 4 GB overlap: $0.088 to $0.20 / hour and $0.085 to $0.34 / hour, since CPU is billed only while busy.** Here is where they differ, from each one's own pages, plus a third option if neither fits.

## The short answer

Tensorlake runs sandboxes for agents as Firecracker microVMs with memory snapshots, and also runs GitHub Actions jobs. Vercel Sandbox runs untrusted code in Firecracker microVMs, billed on active CPU and provisioned memory.

### Pick Tensorlake if you want

- a lower price while the CPU is busy: $0.20 against $0.34 an hour
- pause keeps memory: Yes, memory and running processes (Vercel Sandbox: No)
- CI runners too: Yes, GitHub Actions runners (Vercel Sandbox: No)
- fork and snapshots: Memory snapshots and warm copies (Vercel Sandbox: Snapshots)
- longest session: No limit on Pro; 24 hours on credits, 2 hours free (Vercel Sandbox: 24 hours on Pro, 45 minutes on Hobby)
- self-host: In your VPC or on-prem, on Enterprise, which Vercel Sandbox's pages do not state
- GPUs: H100, by request, which Vercel Sandbox's pages do not state

### Pick Vercel Sandbox if you want

- largest size: 32 vCPU; 8 on Pro (Tensorlake: 16 vCPU on Pro)

### Or Groundbox if you want

- a lower list price than both: $0.08 / hour

Where Tensorlake or Vercel Sandbox is ahead of Groundbox: longest session, self-host, GPUs, largest size, security report.

## Side by side

| Feature | Tensorlake | Vercel Sandbox | Groundbox |
| --- | --- | --- | --- |
| Price, 2 vCPU · 4 GB | [$0.088 to $0.20 / hour](https://docs.tensorlake.ai/platform/billing.md) | [$0.085 to $0.34 / hour](https://vercel.com/docs/vercel-sandbox/pricing) | $0.08 / hour |
| Pause keeps memory | [Yes, memory and running processes](https://docs.tensorlake.ai/sandboxes/lifecycle.md) | [No: the filesystem is saved, not memory](https://vercel.com/docs/sandbox/concepts/persistent-sandboxes) | Yes, memory and disk |
| CI runners too | [Yes, GitHub Actions runners](https://docs.tensorlake.ai/github-actions/quickstart.md) | [No: the pricing page lists sandboxes only](https://vercel.com/docs/vercel-sandbox/pricing) | Yes, on the same key and bill |
| Isolation | [Firecracker microVM](https://www.tensorlake.ai/pricing) | [Firecracker microVM](https://vercel.com/docs/sandbox/concepts/security) | Firecracker VM |
| Fork and snapshots | [Memory snapshots and warm copies](https://docs.tensorlake.ai/sandboxes/snapshots.md) | [Snapshots](https://vercel.com/docs/vercel-sandbox/pricing) | Fork a running sandbox; save points |
| Longest session | [No limit on Pro; 24 hours on credits, 2 hours free](https://docs.tensorlake.ai/sandboxes/lifecycle.md) | [24 hours on Pro, 45 minutes on Hobby](https://vercel.com/docs/vercel-sandbox/pricing) | 720 hours, renewable |
| Self-host | [In your VPC or on-prem, on Enterprise](https://www.tensorlake.ai/pricing) | Not published | No |
| GPUs | [H100, by request](https://docs.tensorlake.ai/sandboxes/gpu.md) | Not published | Not offered |
| Largest size | [16 vCPU on Pro](https://docs.tensorlake.ai/sandboxes/lifecycle.md) | [32 vCPU; 8 on Pro](https://vercel.com/docs/vercel-sandbox/pricing) | 16 vCPU |
| SDK languages | [Python and TypeScript](https://docs.tensorlake.ai/sandboxes/quickstart.md) | [JavaScript / TypeScript and Python](https://vercel.com/docs/vercel-sandbox) | TypeScript and Python |
| Security report | [SOC 2 Type 2; HIPAA on Pro](https://www.tensorlake.ai/pricing) | [SOC 2 Type II](https://vercel.com/docs/sandbox/concepts/security) | No SOC 2 report |
| Free to start | [1 sandbox, 2-hour sessions, no card](https://www.tensorlake.ai/pricing) | [Hobby, within monthly quotas](https://vercel.com/docs/vercel-sandbox/pricing) | $5 of usage, no card |
| Plans | [Free; prepaid usage credits from $5; Pro $250 / month](https://www.tensorlake.ai/pricing) | [Hobby free within monthly quotas; Pro usage comes out of the plan's $20 monthly credit, then rates apply](https://vercel.com/docs/vercel-sandbox/pricing) | Free: everything, with $5 of usage. $20 / month: $20 of usage included, then pay as you go |
| CPU | [$0.07 per active core-hour on credits, $0.042 on Pro; at least 20% of the cores is billed](https://docs.tensorlake.ai/platform/billing.md) | [$0.128 per active CPU-hour (iad1); time waiting on I/O is not billed](https://vercel.com/docs/vercel-sandbox/pricing) | $0.04 per vCPU-hour, a dedicated core |
| Memory | [$0.015 per GB-hour on credits, $0.009 on Pro, billed apart from CPU](https://docs.tensorlake.ai/platform/billing.md) | [$0.0212 per GB-hour of provisioned memory (iad1)](https://vercel.com/docs/vercel-sandbox/pricing) | Included, up to 4 GB per vCPU |
| Billing | [Per second](https://docs.tensorlake.ai/platform/billing.md) | [Memory in 1-minute minimum increments; $0.60 per million sandbox creations](https://vercel.com/docs/vercel-sandbox/pricing) | Per second, no minimum |
| Sizes | [Up to 4 vCPU on credits and 16 on Pro, 1 to 8 GB per vCPU](https://docs.tensorlake.ai/sandboxes/lifecycle.md) | [1 to 32 vCPU, 2 GB per vCPU; 8 vCPU at most on Pro](https://vercel.com/docs/vercel-sandbox/pricing) | 1 to 16 vCPU, up to 4 GB per vCPU |
| Sandboxes at once | [1 free, 100 on credits, 1,000 on Pro](https://www.tensorlake.ai/pricing) | [10 on Hobby, 10,000 on Pro](https://vercel.com/docs/vercel-sandbox/pricing) | 20 once a card is on file |
| Egress controls | [Allow and deny lists by domain, IP or CIDR, changeable while running](https://docs.tensorlake.ai/sandboxes/networking.md) | [Allow all, deny all, or a domain and CIDR allowlist](https://vercel.com/docs/sandbox/concepts/firewall) | An allowlist of hosts per sandbox |

a rust mark means that side is ahead · each value links to its source · checked 2026-11-20

Each against Groundbox: [Groundbox vs Tensorlake](https://groundbox.dev/compare/tensorlake) and [Groundbox vs Vercel Sandbox](https://groundbox.dev/compare/vercel-sandbox). Every other comparison is on [/compare](https://groundbox.dev/compare).

## Sources

### Tensorlake

- [Tensorlake pricing](https://www.tensorlake.ai/pricing), checked 2026-11-20
- [Tensorlake docs](https://docs.tensorlake.ai/), checked 2026-11-20
- [Tensorlake docs: billing](https://docs.tensorlake.ai/platform/billing.md), checked 2026-11-20
- [Tensorlake docs: sandbox lifecycle](https://docs.tensorlake.ai/sandboxes/lifecycle.md), checked 2026-11-20
- [Tensorlake docs: snapshots](https://docs.tensorlake.ai/sandboxes/snapshots.md), checked 2026-11-20
- [Tensorlake docs: networking](https://docs.tensorlake.ai/sandboxes/networking.md), checked 2026-11-20
- [Tensorlake docs: GPU sandboxes](https://docs.tensorlake.ai/sandboxes/gpu.md), checked 2026-11-20
- [Tensorlake docs: quickstart](https://docs.tensorlake.ai/sandboxes/quickstart.md), checked 2026-11-20
- [Tensorlake docs: GitHub Actions](https://docs.tensorlake.ai/github-actions/quickstart.md), checked 2026-11-20

### Vercel Sandbox

- [Vercel Sandbox pricing](https://vercel.com/docs/vercel-sandbox/pricing), checked 2026-11-20
- [Vercel Sandbox docs](https://vercel.com/docs/vercel-sandbox), checked 2026-11-19
- [Vercel Sandbox: security](https://vercel.com/docs/sandbox/concepts/security), checked 2026-11-19
- [Vercel Sandbox: persistent sandboxes](https://vercel.com/docs/sandbox/concepts/persistent-sandboxes), checked 2026-11-20
- [Vercel Sandbox: firewall](https://vercel.com/docs/sandbox/concepts/firewall), checked 2026-11-19

## Try the third option free.

[Start free, no card](https://groundbox.dev/signup)[Groundbox vs Tensorlake](https://groundbox.dev/compare/tensorlake)
