Vercel Sandbox vs Modal VM Sandboxes

Both run each sandbox in its own VM with its own kernel, and both stop a session within 24 hours. Vercel Sandbox is a Firecracker microVM that saves its files on every stop and runs in 19 regions, while Modal VM Sandboxes run on Cloud Hypervisor with systemd and nested virtualization, next to Modal's inference and training.

The companies and products

Vercel Sandbox

Vercel is a cloud platform for web apps and agents, and Vercel Sandbox is its product for running untrusted or agent-written code in isolated Linux microVMs.

Each sandbox is a Firecracker microVM with its own kernel, started from a Vercel image, your own OCI image, or a saved snapshot. You drive it with the JavaScript or Python SDK or the sandbox CLI: run commands, move files, open ports, and set network rules. It is aimed at code interpreters, app builders, test runs and agent tools, and Vercel's docs say it is not meant to run nonstop.

Modal VM Sandboxes

Modal is a serverless cloud for AI work, and VM Sandboxes are the version of its sandboxes that run in a virtual machine with their own Linux kernel.

You create a VM Sandbox from Python, TypeScript or Go code with runtime="vm", run commands in it, and terminate it when the job is done. Modal built the runtime on the Rust-based Cloud Hypervisor, so Docker, FUSE, systemd and nested cgroups work, with the same API, images and prices as its gVisor sandboxes. It became generally available on 1 October 2026 and suits teams that want many short-lived agent machines next to Modal's inference and training.

Which use cases each one fits

For each use case, Battleships prices every sandbox provider on the same workload and sorts them from the cheapest monthly bill up. “22nd cheapest of 345” means 21 of the other 344 providers run that workload, with every feature it needs, for less. “With limits” means the provider runs it but misses something, so it is listed after every full match. Open a use case for its whole ranking and what each provider misses.

Use case and its workloadVercel Sandbox: price rankModal VM Sandboxes: price rankbochka: price rank
Software factory
100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month
Neither has a desktop to check the app an agent built, and both stop at 24 hours; Modal VM Sandboxes rank 84th, Vercel Sandbox 244th, and bochka ranks 3rd.
244th of 367, with limits
$3,351 a month
84th of 367, with limits
$759 a month
3rd cheapest of 367
$634 a month
Always-on AI employees
50 machines 24/7 · 2 vCPU / 4 GB
Both stop each session at 24 hours, so an agent that lives on its machine all month fits neither; bochka has no session cap on paid plans.
257th of 366, with limits
$3,657 a month
Can't run it5th cheapest of 366
$657 a month
Code interpreter bursts
100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month
Short code bursts fit Vercel Sandbox well because it bills CPU only while code uses it; bochka's start limits slow very large bursts.
22nd cheapest of 345
$216 a month
Can't run it186th of 345, with limits
$2,000 a month
Sandboxes for your users
300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month
Vercel Sandbox ranks 15th for sandboxes you hand to your users, and bochka ranks 1st for the same work.
15th cheapest of 361
$1,267 a month
Can't run itCheapest of 361
$209 a month
RL rollouts
1,000 at once · 2 vCPU / 4 GB · 1,000,000 sessions of 3 min a month
Huge bursts of short rollouts suit Vercel Sandbox best here: it ranks 3rd, against 171st for Modal VM Sandboxes and 34th for bochka.
3rd cheapest of 360
$15,761 a month
171st of 360, with limits
$16,322 a month
34th of 360, with limits
$2,000 a month
Agent evals
250 at once · 4 vCPU / 8 GB · 15,000 sessions of 20 min a month
69th cheapest of 364
$1,508 a month
Can't run it10th cheapest of 364
$200 a month
CI runners
40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month
Both run Docker builds; Vercel Sandbox ranks 107th and Modal VM Sandboxes 188th, while bochka ranks 4th with /dev/kvm on every plan.
107th cheapest of 377
$1,144 a month
188th of 377, with limits
$1,064 a month
4th cheapest of 377
$86 a month
24/7 dev boxes
1 machine 24/7 · 4 vCPU / 8 GB
276th of 368, with limits
$145 a month
Can't run it18th cheapest of 368
$26 a month
Computer-use agents
50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month
Neither ships a desktop or browser, so you build one yourself; bochka streams a 60fps desktop with a computer-use API.
235th of 369, with limits
$952 a month
Can't run it6th cheapest of 369
$200 a month
Browser-use agents
50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month
259th of 367, with limits
$486 a month
Can't run it7th cheapest of 367
$200 a month
Scraping & bots (startup scale)
50 at once · 1 vCPU / 2 GB · 100,000 sessions of 3 min a month
277th of 346, with limits
$468 a month
Can't run it6th cheapest of 346
$90 a month
Mass scraping & bots (undetectable)
1,000 at once · 1 vCPU / 2 GB · 3,000,000 sessions of 2 min a month
234th of 338, with limits
$9,960 a month
Can't run it71st of 338, with limits
$1,800 a month
Parallel red teaming
200 at once · 2 vCPU / 4 GB · 5,000 sessions of 30 min a month
Both are a compromise for many parallel attack machines, at 283rd for Vercel and 294th for Modal, while bochka ranks 2nd.
283rd of 360, with limits
$554 a month
294th of 360, with limits
$642 a month
2nd cheapest of 360
$200 a month
GPU eval jobs
10 at once · 8 vCPU / 32 GB · 500 sessions of 30 min a month
Can't run itCan't run itCan't run it
macOS agents
5 at once · 4 vCPU / 8 GB · 1,000 sessions of 20 min a month
Can't run itCan't run itCan't run it

Key features

FeatureVercel SandboxModal VM Sandboxesbochka
IsolationFirecracker microVM with its own kernelVM with its own Linux kernel on Cloud HypervisorFull VM with its own kernel on dedicated AMD Ryzen 9 9950X hosts
CPU benchmark (4 vCPU / 8 GB)12.96 runs/s15.54 runs/s26.69 runs/s
Longest session24 hours on Pro and Enterprise, 45 minutes on Hobby24 hours, default 5 minutesNo cap on paid plans; you set the time to live
Saving filesFilesystem snapshot on every stop, kept 30 days after last use; the sandbox resumes from itNo paused state; only Volumes and snapshots you take keep filesDisk saved every minute and on stop; a stopped sandbox costs nothing
Memory snapshotsNo; processes start over on every resumeYes, in alpha: snapshotting ends the sandbox and snapshots expire after 7 daysNo; resume boots the VM again with its disk
Docker and systemdDocker with sudoDocker, systemd and FUSE; start dockerd yourselfDocker, root and systemd in every sandbox
Nested virtualizationNoTeam and Enterprise plans only, on requestYes, /dev/kvm on every plan
Desktop, browser, computer useNone; headless Linux onlyNot offered; an example shows how to set up Xvfb and VNC yourself60fps desktop, Chromium, computer-use and browser-control APIs
Coding agentsClaude Code, Codex, OpenCode and pi preinstalled; you start them as commandsExamples only; you install the agent yourselfClaude Code, Codex and others preinstalled, with an API to prompt, steer and stream events
Keys kept outside the sandboxCredentials brokering injects headers for any HTTPS domain you allowSecrets are environment variables, or you run your own proxy in a sidecarCredentials proxy for model providers and any HTTPS API; switch to another account when one runs out
Inbound accessPublic URLs for up to 15 ports, no built-in auth; no raw TCPEncrypted and raw TCP tunnels, connect tokens, inbound IP allowlistHTTPS URLs for up to 50 ports, token-protected by default, plus raw TCP
Outbound network rulesDeny-all or a domain and CIDR allowlist, changeable while runningBlock all, IP-range allowlist, domain allowlist (beta)Open internet, no provider-side allowlist; firewall inside the VM
Custom imagesYour own OCI images from a Dockerfile, via Vercel Container RegistryCustom images through the same API as Modal's other sandboxesNamed snapshots (templates), shareable by code; no Dockerfile builds
Regions19 regions across the US, Europe, Asia, South America, Africa and AustraliaUS, EU and Asia-Pacific; pinning a region costs 1.15x to 1.75xEU only: Germany, Finland, France
Every feature Battleships tracks (43)
FeatureVercel SandboxModal VM Sandboxesbochka
Full VM (own kernel)YesYesYes
Docker insideYesYesYes
Nested virtualizationNoYesYes
RootYesYesYes
SystemdNot publishedYesYes
Desktop GUINoNoYes
BrowserNoNoYes
Computer-use APINoNoYes
SSHNoYesYes
Persistent diskYesYesYes
Pause/resumeYesNoYes
Automatic snapshotsYesNoYes
Fork/cloneYesYesYes
Live fork (no pause)YesYesYes
Memory forkNoYesNo
Custom image (Docker or snapshot)YesYesYes
≥24 h sessionsYesYesYes
Idle auto-stopNoYesNo
Wake on requestNoNoNo
HTTPS preview URLsYesYesYes
Raw TCP inboundNoYesYes
Public IPv4NoNoNo
Egress allowlistYesYesNo
Secret proxyYesNoYes
Secret proxy for any APIYesNoYes
Preinstalled agentsYesNoYes
Agent harness APINoNoYes
MCP serverYesNoNo
GPUNoNoNo
WindowsNoNoNo
Arm64NoNoNo
Self-hostingNoNoNo
BYOCNoNoNo
Open sourceNoNoNo
Scoped API keysNot publishedNot publishedYes
Spend limitsYesNot publishedYes
Usage per sandbox (API)NoNot publishedYes
Audit logsNot publishedYesNo
SSOYesYesNo
SOC 2YesYesNo
HIPAAYesYesNo
EU data residencyYesYesYes
Zero data retentionNot publishedNot publishedYes

Performance

Vercel Sandbox12.96 runs/s
Modal VM Sandboxes15.54 runs/s
bochka26.69 runs/s

All three ran the same hpc-sandbox-benchmarks Node.js build-and-test loop at 4 vCPU / 8 GB on 29 September 2026. Modal VM Sandboxes did 15.54 runs per second and Vercel Sandbox did 12.96, so Modal did about 1.2 times the CPU work. bochka did 26.69, about 1.7 times Modal and about twice Vercel; this measures CPU speed only, not start time.

Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. Modal VM Sandboxes measured on: VM Sandbox, request = size.

Pricing in five scenarios

Each bill is priced from the providers' public price lists by the open Battleships engine, with the cheapest plan that runs the scenario.

ScenarioVercel Sandbox, a monthModal VM Sandboxes, a monthbochka, a month
One coding agent, workdays
4 vCPU / 8 GB, 50 GB disk, 8 hours a day, 22 days a month, files kept between days
$57 · Pro$15 · Starter · with limits$20 · $20/mo
One agent that never stops
4 vCPU / 8 GB, 50 GB disk, running 24/7 all month
$236 · Pro · with limits$178 · Starter · with limits$26 · $20/mo
50 agents in parallel
50 machines of 4 vCPU / 8 GB at once, 8-hour sessions, 1,100 sessions a month, Docker inside
$2,884 · Pro$2,229 · Starter · with limits$317 · $20/mo × 3 seats
Sandboxes for your users
2 vCPU / 4 GB, 30,000 sessions of 10 minutes a month, up to 300 at once
$828 · Pro$792 · Team$200 · $100/mo × 2 seats
1,000 short jobs at once
2 vCPU / 4 GB, 1,000 at once, 3 minutes each, 200,000 jobs a month
$3,152 · Pro$3,193 · Team · with limits$2,000 · $2000/mo · with limits

Vercel Sandbox: Vercel charges for CPU only while your code uses it, but for memory the whole time the sandbox runs, plus a small fee per sandbox created and a monthly fee per GB of snapshots. The $20 Pro credit is shared with your other Vercel usage.

Modal VM Sandboxes: Modal bills by the second for whichever is higher, the CPU and memory you request or what you use, until the sandbox ends. Starter is free with $30 of monthly credit, Team is $250 a month with $100 of credit, and pinning a region multiplies the price.

Modal VM Sandboxes: bochka charges one hourly rate per machine size, by the second, only while the sandbox runs. Plans start at $20 a month and the whole fee becomes usage credit; a stopped sandbox costs nothing and keeps its disk.

Limits in those scenarios

For one agent that never stops, both end each session at 24 hours: Vercel resumes from its snapshot with processes restarted, and Modal keeps only what is in a Volume or a snapshot. That Volume rule is also why Modal does not fit the workday and 50-agent scenarios, where installed packages reset. For 1,000 short jobs at once, Vercel fits, while Modal's Team price uses shared vCPUs that other tenants can slow, and bochka's $2,000 plan hits its start limits of 600 an hour and 2,400 a day.

ScenarioVercel SandboxModal VM Sandboxesbochka
One coding agent, workdaysNone that bite
  • files survive a stop only in Volumes: packages installed elsewhere reset
None that bite
One agent that never stops
  • Pro: sessions capped at 24 h
  • Starter: sessions capped at 24 h
  • files survive a stop only in Volumes: packages installed elsewhere reset
None that bite
50 agents in parallelNone that bite
  • files survive a stop only in Volumes: packages installed elsewhere reset
None that bite
Sandboxes for your usersNone that biteNone that biteNone that bite
1,000 short jobs at onceNone that bite
  • shared vCPUs on this tier (Modal (VM Sandboxes) sells dedicated cores separately as “VM Sandbox, request = size”): heavy, sustained CPU use can be slowed by other tenants
  • max 2,400 starts/day (needs ~6,667)
  • max 600 starts/hour (needs ~1,000)
  • bursts above 90 starts/min are queued (a burst of 333 takes ~4 min to start)

Which one should you use?

Use Vercel Sandbox when

you run many short bursts of untrusted code close to users around the world, want files saved on every stop without extra code, or already deploy on Vercel.

  • Regions all over the world. Useful if your users are spread out: Vercel Sandbox runs in 19 regions, from Virginia to Tokyo, Sydney and Cape Town.
  • Files come back on their own. Every stop saves a filesystem snapshot and the next session resumes from it, with no Volume code to write.
  • Dockerfile images and brokered keys. Useful if you already have a Dockerfile: boot sandboxes straight into it, and inject API keys for allowed domains from outside.

Use Modal VM Sandboxes when

you need systemd, raw TCP or memory snapshots in short-lived VMs, or already run your models and GPUs on Modal.

  • A fuller Linux kernel. Useful for agents that need systemd, FUSE, raw TCP or SSH, and nested virtualization on Team and Enterprise.
  • Memory snapshots and big bursts. Modal can clone a sandbox with its processes still running (alpha) and runs 5,000 containers at once on Team.
  • One platform with inference and training. Useful if you already serve or train models on Modal: sandboxes and models sit on one account and one bill.

Where bochka fits

Vercel Sandbox and Modal VM Sandboxes are both built for short jobs that end within a day. If your agents work for hours or days on one machine, with builds, Docker, emulators and a browser to check their work, look at bochka. It gives each agent a full VM with no session cap on paid plans, on a faster CPU.

Agents that run past a day

Both stop each session at 24 hours, so a long agent task is cut off. Vercel resumes from a snapshot but restarts every process; on Modal you write snapshot or Volume code yourself. bochka has no session cap and saves its disk every minute.

A desktop and a browser built in

Neither has a screen, so an agent cannot open the app it built to see if it works. On Modal you set up Xvfb and VNC yourself, and on Vercel you build it from scratch. bochka streams a 60fps desktop with Chromium and computer-use APIs.

Faster CPU and /dev/kvm everywhere

Installs, builds and tests are billed while they wait. bochka did 26.69 runs/s against 15.54 for Modal and 12.96 for Vercel, and gives /dev/kvm on every plan; Vercel has no nested virtualization and Modal only on request.

Use bochka when your agents work for hours or days on one machine: builds, Docker, emulators and a desktop with a browser, on a faster VM that keeps its disk, with an API to drive the agents inside from outside. Compare it one to one: bochka vs Vercel Sandbox and bochka vs Modal VM Sandboxes.

Questions people ask

Is Vercel Sandbox cheaper than Modal VM Sandboxes?

Usually not. Modal costs less in four of our five scenarios, for example $178 against $236 for one agent that never stops and $2,229 against $2,884 for 50 parallel agents, but in three of those its files reset outside a Volume. For 1,000 short jobs at once Vercel is a little cheaper at $3,152 against $3,193, and bochka costs $26 a month for the always-on agent.

Which is faster, Vercel Sandbox or Modal VM Sandboxes?

On the hpc-sandbox-benchmarks CPU test at 4 vCPU / 8 GB, Modal VM Sandboxes did 15.54 runs per second and Vercel Sandbox did 12.96. bochka did 26.69 in the same run. This measures CPU speed only, not start time.

Can Vercel Sandbox and Modal VM Sandboxes run Docker?

Yes, both run Docker inside their VM: Vercel with sudo, and Modal with systemd, where you start dockerd yourself. Vercel has no nested virtualization, and Modal offers it on Team and Enterprise after you ask support. bochka runs Docker and exposes /dev/kvm on every plan.

Can either run 24/7?

Not as one session. Vercel Sandbox stops after 24 hours on Pro and Enterprise and resumes from its filesystem snapshot with processes restarted. Modal VM Sandboxes also end within 24 hours, and only Volumes and snapshots keep files. bochka has no session cap on paid plans.

Do they have a desktop or browser for computer use?

No. Vercel Sandbox is headless Linux only, and Modal shows an example of setting up Xvfb and VNC yourself. bochka includes a 60fps desktop, Chromium, and computer-use and browser-control APIs.

Try bochka

Install the CLI, then run bochka new.

Product preview

More comparisons

Bochka product preview · Pricing, benchmarks and commands are illustrative. Sandbox provisioning is not connected.