bochka vs Notte

Notte is a managed cloud browser: your agent gets Chrome through an API, with residential proxies and captcha solving turned on. bochka gives the agent a whole VM with its own kernel, a desktop and Chrome, Docker and a disk that stays between runs.

Where bochka stands out

A whole computer, not only a browser

A bochka sandbox is a full VM with its own kernel, root, systemd, Docker and /dev/kvm, and Chrome runs on it. The agent can download a file, process it with a script, start a database and check the result in the browser, all on one machine. Notte gives you a browser to drive and Python Functions, with no shell.

Nothing resets after a few hours

Paid bochka plans have no session cap. The disk persists, snapshots are taken every minute and on stop, a stopped sandbox costs nothing, and you can fork a running sandbox with its Chrome profiles. Notte sessions end after 15, 120 or 300 minutes by plan and start with a fresh browser unless you save a profile.

A desktop and coding agents built in

bochka streams a 60fps Linux desktop and Chrome that you can embed and record, and every built-in agent, such as Claude Code or Codex, can see and use it. You drive those agents from outside with a prompt, steer and events API. The credentials proxy keeps real AI keys out of the sandbox for any HTTPS API.

The companies and products

bochka

bochka.xyz runs sandboxes for AI agents: each one is a full Linux VM with its own kernel.

You start a machine with one command or one API call, then SSH in, run Docker, open a desktop, or let a built-in agent (Claude Code, Codex, OpenCode and others) work in it. The disk stays between sessions, and a stopped sandbox costs nothing.

Notte

Notte is a cloud browser platform for AI agents: you start hosted Chrome sessions through an API, SDK, CLI or MCP server.

Each Notte session is a hosted browser you control over CDP, Playwright or Notte's own actions, with residential proxies, captcha solving and a consistent fingerprint on by default. Around it Notte sells a task-driven browser agent, vaults for passwords, personas with their own email and phone number, and Python Functions that turn a browser script into an API endpoint. It is a different kind of product from bochka: a browser service, not a general sandbox, so this page compares the two where they overlap, which is agents that use the web.

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. “3rd cheapest of 367” means 2 of the other 366 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 workloadbochka: price rankNotte: price rank
Software factory
100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month
Notte has no shell, Docker or lasting disk, so a coding agent cannot build and test there; it can only use Notte as its browser.
3rd cheapest of 367
$634 a month
44th of 367, with limits
$121 a month
Always-on AI employees
50 machines 24/7 · 2 vCPU / 4 GB
Notte sessions end after at most 300 minutes on Startup and start fresh, so an always-on assistant needs a machine like bochka to live on.
5th cheapest of 366
$657 a month
99th of 366, with limits
$365 a month
Code interpreter bursts
100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month
186th of 345, with limits
$2,000 a month
220th of 345, with limits
$100 a month
Sandboxes for your users
300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month
Cheapest of 361
$209 a month
195th of 361, with limits
$105 a month
RL rollouts
1,000 at once · 2 vCPU / 4 GB · 1,000,000 sessions of 3 min a month
34th of 360, with limits
$2,000 a month
222nd of 360, with limits
$500 a month
Agent evals
250 at once · 4 vCPU / 8 GB · 15,000 sessions of 20 min a month
10th cheapest of 364
$200 a month
227th of 364, with limits
$100 a month
CI runners
40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month
4th cheapest of 377
$86 a month
230th of 377, with limits
$100 a month
24/7 dev boxes
1 machine 24/7 · 4 vCPU / 8 GB
Notte has no shell or disk to develop on, so it does not fit this use.
18th cheapest of 368
$26 a month
255th of 368, with limits
$100 a month
Computer-use agents
50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month
Notte controls a browser only, so agents that need a whole desktop or other apps fit bochka's 60fps Linux desktop better.
6th cheapest of 369
$200 a month
76th of 369, with limits
$100 a month
Browser-use agents
50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month
Both fit: Notte is built for this with stealth and captcha solving on by default, and bochka gives the agent Chrome on a full VM next to its code and files.
7th cheapest of 367
$200 a month
2nd cheapest of 367
$100 a month
Scraping & bots (startup scale)
50 at once · 1 vCPU / 2 GB · 100,000 sessions of 3 min a month
Both fit; Notte routes traffic through residential proxies billed per GB, while bochka uses datacenter IPs and lets you bring your own proxy.
6th cheapest of 346
$90 a month
7th cheapest of 346
$100 a month
Mass scraping & bots (undetectable)
1,000 at once · 1 vCPU / 2 GB · 3,000,000 sessions of 2 min a month
Notte is the stronger fit for sites that block bots at scale, since residential proxies, fingerprints and captcha solving come built in; on bochka you add those yourself.
71st of 338, with limits
$1,800 a month
2nd of 338, with limits
$975 a month
Parallel red teaming
200 at once · 2 vCPU / 4 GB · 5,000 sessions of 30 min a month
2nd cheapest of 360
$200 a month
143rd of 360, with limits
$100 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 it
macOS agents
5 at once · 4 vCPU / 8 GB · 1,000 sessions of 20 min a month
Can't run itCan't run it

Key features

FeaturebochkaNotte
What you getA full Linux VM with its own kernel, with Chrome and a desktop on itA hosted browser session you reach through the API; isolation model not published
Browser controlChrome in the VM: Playwright, Puppeteer or CDP on a local port, plus a 1920x1080 60fps browser streamHosted Chromium or Chrome with a CDP URL, Playwright, Notte actions and a live view you can embed
Getting past bot checksDatacenter IPs; bring your own proxy or captcha serviceResidential proxies ($10/GB, on by default), captcha solving and a matching Chrome fingerprint on every plan
How long a session runsNo session cap on paid plans15 min on Free, 120 min on Developer, 300 min on Startup
What survives a stopThe whole disk, with snapshots every minute and on stop; Chrome profiles go with snapshots and forksFresh browser each session; an opt-in profile keeps cookies, local storage and session storage
ForkFork a running sandbox, browser and files includedNo session fork; a Function can be forked as a copy of its code
Running your own codeRoot shell, any language, Docker, systemd and /dev/kvmPython Functions with a run() entry point, called by API, webhook or schedule; no shell
Desktop and computer use60fps Linux desktop that every built-in agent can see and use, embeddable and recordable to MP4Browser only; no desktop
AgentsClaude Code, Codex and other coding agents inside the VM, driven from outside with a prompt, steer and events APINotte's own browser agent: start it with a task, stream its logs, stop it, export the run as code; an MCP server for agents outside
Keeping secrets from the modelAI credentials proxy: real keys never enter the sandbox, for model providers and any HTTPS APIVaults fill passwords, cards and 2FA codes so the LLM never sees them; personas with their own email and phone
Reaching it from outsideHTTPS URLs for up to 50 ports, token-protected; raw TCP; dedicated IPv6 or IPv4No open ports; a short-lived signed live-view URL per session
Self-hostingHosted onlySource-available framework runs on local Chromium; on-premise and bring-your-own-cloud on Enterprise
Where it runsEU only (Germany, Finland, France), with EU data residency and zero data retentionSession region not published; built-in proxies exit in the US unless you pick a country
Compliance and transparencySOC 2 is in progress. We're open about where we stand: live status and data retention policy.SOC 2
Every feature Battleships tracks (43)
FeaturebochkaNotte
Full VM (own kernel)YesNo
Docker insideYesNo
Nested virtualizationYesNo
RootYesNot published
SystemdYesNot published
Desktop GUIYesNo
BrowserYesYes
Computer-use APIYesYes
SSHYesNo
Persistent diskYesNo
Pause/resumeYesNo
Automatic snapshotsYesNo
Fork/cloneYesNo
Live fork (no pause)YesNo
Memory forkNoNo
Custom image (Docker or snapshot)YesNot published
≥24 h sessionsYesYes
Idle auto-stopNoYes
Wake on requestNoNo
HTTPS preview URLsYesNo
Raw TCP inboundYesNot published
Public IPv4NoNo
Egress allowlistNoNo
Secret proxyYesNo
Secret proxy for any APIYesNo
Preinstalled agentsYesNo
Agent harness APIYesNo
MCP serverNoYes
GPUNoNo
WindowsNoNo
Arm64NoNot published
Self-hostingNoYes
BYOCNoYes
Open sourceNoYes
Scoped API keysYesNot published
Spend limitsYesNot published
Usage per sandbox (API)YesNo
Audit logsNoNot published
SSONoYes
SOC 2NoYes
HIPAANoYes
EU data residencyYesNot published
Zero data retentionYesNot published

Performance

bochka26.69 runs/s
NotteNot measured

There is no comparable measurement. Notte does not publish the machine size behind a browser session and was not in the StarSling benchmark run, where bochka did 26.69 Node.js build-and-test runs per second at 4 vCPU / 8 GB. For browsing work, speed depends mostly on the sites, the proxy and the model.

Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. Notte has no published run in this suite yet, so there is no like-for-like number to compare.

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.

Scenariobochka, a monthNotte, 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
$20 · $20/mo$100 · Startup · with limits
One agent that never stops
4 vCPU / 8 GB, 50 GB disk, running 24/7 all month
$26 · $20/mo$100 · Startup · with limits
50 agents in parallel
50 machines of 4 vCPU / 8 GB at once, 8-hour sessions, 1,100 sessions a month, Docker inside
$317 · $20/mo × 3 seats$100 · Startup · with limits
Sandboxes for your users
2 vCPU / 4 GB, 30,000 sessions of 10 minutes a month, up to 300 at once
$200 · $100/mo × 2 seats$100 · Startup · with limits
1,000 short jobs at once
2 vCPU / 4 GB, 1,000 at once, 3 minutes each, 200,000 jobs a month
$2,000 · $2000/mo · with limits$100 · Startup · with limits

bochka: You pick a monthly plan from $20, and all of it becomes usage credit spent per second on running sandboxes. A stopped sandbox costs nothing and keeps its disk.

Notte: Notte's plan fee comes back as credits that expire each cycle, spent on browser sessions ($0.01 an hour), Functions ($0.05 an hour), residential proxies ($10 per GB) and model tokens at cost. Proxies are on by default, so a session's traffic is billed per GB unless you turn them off or bring your own proxy on Startup.

Limits in those scenarios

Notte fits no scenario in the table, because each needs a machine: work that runs for more than 300 minutes, files that stay, or Docker. Notte sessions end at the plan's time limit and start with a fresh browser, and Startup allows 100 browsers at once, so 300 sandboxes for your users or 1,000 jobs at once need Enterprise. On bochka the limit for the burst is starts: the $2,000 plan allows 2,400 starts a day and queues bursts above 90 starts a minute.

ScenariobochkaNotte
One coding agent, workdaysNone that bite
  • Startup: sessions capped at 5 h
  • missing persistent disk
  • machine size not published
  • browser-session product, not a general sandbox
One agent that never stopsNone that bite
  • Startup: sessions capped at 5 h
  • missing persistent disk
  • machine size not published
  • browser-session product, not a general sandbox
50 agents in parallelNone that bite
  • Startup: sessions capped at 5 h
  • missing persistent disk, Docker inside
  • machine size not published
  • browser-session product, not a general sandbox
Sandboxes for your usersNone that bite
  • Startup: max 100 concurrent
  • machine size not published
  • browser-session product, not a general sandbox
  • different kind of product: browser session
1,000 short jobs at once
  • 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)
  • Startup: max 100 concurrent
  • machine size not published
  • browser-session product, not a general sandbox
  • different kind of product: browser session

Which one should you use?

Use Notte when

the job is pure browsing on sites that fight bots, and you want proxies, captcha solving, a browser agent and identities handled for you.

  • Getting past bot checks. Useful if your agents browse sites that block bots: Notte turns on residential proxies, captcha solving and a matching fingerprint by default.
  • Ready-made browsing tools. Useful if you want browsing handled for you: a task-driven browser agent, personas with email and phone, and scheduled browser Functions.
  • Open source and on-premise. Useful if you must run it yourself: the framework is source-available, and Enterprise adds on-premise and bring-your-own-cloud.

Use bochka when

the agent does more than browse: it writes and runs code, keeps files and Chrome logins between runs, works for hours, or needs a desktop, Docker and coding agents on one VM.

  • A whole computer, not only a browser. On Notte, anything that is not browsing has to run somewhere else, so you run a second machine and move files between it and the browser session.
  • Nothing resets after a few hours. A long task on Notte has to fit in 300 minutes on Startup, and only cookies and storage carry over, so downloads and work files are gone when the session closes.
  • A desktop and coding agents built in. Notte controls a browser only, so a task that also needs a terminal or a desktop app needs another tool, and its built-in agent browses rather than writes code.

Questions people ask

Is bochka cheaper than Notte?

Notte's plan fee is lower in most of our scenarios: we priced every one at the $100 a month Startup plan, against $20 to $2,000 on bochka. But none of the five fit on Notte, because they need sessions longer than 300 minutes, a disk or Docker. For browser-use agents, where both fit, Notte comes to $100 a month in our model and bochka to $200.

Is Notte a sandbox?

Not in the usual sense. Notte hosts browser sessions and runs Python Functions, but it gives you no shell, no Docker and no disk that stays. bochka gives every agent a full Linux VM with Chrome on it.

Can bochka solve captchas or use residential proxies?

Not out of the box. bochka sandboxes use datacenter IPs and have no built-in captcha solver, but they have open internet, so you can point Chrome at your own proxy or captcha service. Notte turns on residential proxies and captcha solving by default.

Can I use Notte from a bochka sandbox?

Yes. A bochka sandbox has open internet, so code or an agent inside it can call Notte's API, SDK or MCP server. That lets the agent build and test on bochka and send hard-to-reach pages to Notte.

Does Notte keep my files between sessions?

No. Each Notte session starts with a fresh browser, and an opt-in profile carries over only cookies, local storage and session storage. On bochka the whole disk persists, with snapshots every minute and on stop.

Is bochka SOC 2 compliant?

SOC 2 is in progress. We're open about where we stand: live status at status.bochka.xyz and our data retention policy at docs.bochka.xyz/data-retention. Notte shows a SOC 2 Type II badge on its site and lists HIPAA for Enterprise.

Start a sandbox

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.