bochka vs GitHub Actions

GitHub Actions gives each CI job a fresh VM that is thrown away when the job ends, at most 6 hours later. bochka gives each agent a full VM that keeps its disk, runs as long as you want and can be reached from outside.

Where bochka stands out

A machine that stays

A bochka sandbox keeps its disk. It is saved every minute and when it stops, a stopped sandbox costs nothing, and paid plans have no session cap. You can also fork a running sandbox to try two ideas from the same state.

Reach it from outside

You can SSH into a bochka sandbox, open HTTPS URLs on up to 50 ports, take raw TCP and watch its desktop at 60fps. An agent can run a dev server or an app and you, a test or a user can use it while it runs.

Agents you drive from outside, keys kept out

Claude Code, Codex and five other agents come preinstalled, and you prompt, steer and follow them through one API from any program, not only from issues and pull requests. Real API keys stay in a credentials proxy outside the sandbox.

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.

Isolation: Full VM per sandbox, own kernel
Hardware: Dedicated AMD Ryzen 9 9950X (Zen 5) hosts
Regions: EU: Germany, Finland, France
Plans: From $20 a month, all of it usage credit, 7-day trial

GitHub Actions

GitHub Actions is GitHub's CI/CD service: workflows in your repository run as jobs on runners that GitHub hosts or that you host yourself.

GitHub Actions is a CI service, not a sandbox API. Each job on a GitHub-hosted runner gets a new VM, Linux and Windows on Azure and macOS on GitHub's own Mac fleet, with passwordless sudo, and the VM is gone when the job ends. It is built for building, testing and deploying the code in a repository, and GitHub's own Copilot cloud agent and the Claude and Codex agents on GitHub also run on Actions runners.

Kind of product: CI/CD service; jobs are defined in workflow files in a GitHub repository
Isolation: A new VM for each job; Linux and Windows on Azure, macOS on GitHub's own Mac cloud
Job length: Each job can run for up to 6 hours
Concurrency: 20 jobs on Free, 40 on Pro, 60 on Team, 500 on Enterprise; macOS 5, or 50 on Enterprise
Allowed use: GitHub-hosted runners are for work on the repository's software project; reselling Actions as a service is not allowed
Self-hosted runners: Open-source runner under MIT; usage on self-hosted runners is free
Compliance: SOC 1 and SOC 2 Type 2 reports, ISO/IEC 27001:2022

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 workload	bochka: price rank	GitHub Actions: price rank
Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month A software factory needs agents that keep their checkout, Docker cache and services between tasks, which a fresh 6-hour CI job does not give you.	3rd cheapest of 367 $634 a month	298th of 367, with limits $5,853 a month
Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB An assistant that lives on its machine all month cannot fit in jobs that end after 6 hours and start from a clean VM.	5th cheapest of 366 $657 a month	333rd of 366, with limits $10,692 a month
Code interpreter bursts 100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month Each job waits for a fresh VM and bills at least a minute, which suits builds better than many tiny code runs.	186th of 345, with limits $2,000 a month	294th of 345, with limits $721 a month
Sandboxes for your users 300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month GitHub's terms do not allow offering Actions to your own users as a service, and its runners take no inbound traffic, so it is not built for this.	Cheapest of 361 $209 a month	308th of 361, with limits $2,892 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	304th of 360, with limits $14,721 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	303rd of 364, with limits $1,962 a month
CI runners 40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month Both fit; GitHub Actions is the native choice for a GitHub repo's CI, while bochka suits CI that needs a warm disk between runs or jobs longer than 6 hours.	4th cheapest of 377 $86 a month	100th cheapest of 377 $1,012 a month
24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB A dev box needs SSH, open ports and a disk that stays; hosted runners have none of these.	18th cheapest of 368 $26 a month	279th of 368, with limits $159 a month
Computer-use agents 50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month Runners are headless, while bochka gives each agent a desktop streamed at 60fps with computer use built in.	6th cheapest of 369 $200 a month	292nd of 369, with limits $1,768 a month
Browser-use agents 50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month	7th cheapest of 367 $200 a month	135th of 367, with limits $942 a month
Scraping & bots (startup scale) 50 at once · 1 vCPU / 2 GB · 100,000 sessions of 3 min a month	6th cheapest of 346 $90 a month	134th of 346, with limits $1,184 a month
Mass scraping & bots (undetectable) 1,000 at once · 1 vCPU / 2 GB · 3,000,000 sessions of 2 min a month	71st of 338, with limits $1,800 a month	301st of 338, with limits $28,971 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	70th of 360, with limits $467 a month
GPU eval jobs 10 at once · 8 vCPU / 32 GB · 500 sessions of 30 min a month	Can't run it	Can't run it
macOS agents 5 at once · 4 vCPU / 8 GB · 1,000 sessions of 20 min a month GitHub Actions has hosted macOS runners on Apple silicon; bochka runs Linux only.	Can't run it	25th of 36, with limits $1,178 a month

Key features

Feature	bochka	GitHub Actions
Kind of product	Sandbox API: create, stop, fork and reach machines from your code	CI service: jobs run from workflow files in a GitHub repository
Isolation	Full VM with its own kernel on dedicated AMD Ryzen 9 9950X hosts, with root and systemd	New VM per job on Azure (Linux, Windows) or GitHub's Mac cloud, with passwordless sudo
Session length	No cap on paid plans; you set the time to live	Up to 6 hours per job
Disk between runs	Disk persists, saved every minute and on stop; a stopped sandbox costs nothing	Every job starts on a fresh VM; files carry over only through cache, artifacts or custom images
Fork	Fork a running sandbox with its disk	No; custom images are a template for new runners, not a copy of a live one
Nested virtualization	Yes, /dev/kvm for Android emulators and VMs	Yes on Linux runners, including the 2-vCPU ones, after adding the runner user to the KVM group
Inbound access	SSH, HTTPS URLs for up to 50 ports, raw TCP, dedicated IPv6 or IPv4	None; GitHub says hosted runners never need inbound connections
Desktop and computer use	Linux desktop streamed at 60fps, Chromium, computer use API	Headless runners; browsers are preinstalled for tests
Coding agents	Seven agents preinstalled in your sandbox, driven from outside with a prompt, steer and events API	Copilot cloud agent, plus Claude and Codex in public preview, working on issues and pull requests on Actions runners
Keeping API keys out of the machine	Credentials proxy for model providers and any HTTPS API; switch a credential to another account when credits run out	Secrets enter the job as inputs or environment variables and are hidden from logs
Operating systems	Linux x64	Linux x64 and arm64, Windows x64 and arm64, macOS
GPUs	None	Linux and Windows larger runners with an NVIDIA T4
Regions and networking	EU only (Germany, Finland, France), open internet	Azure; larger runners can join your Azure network in regions across the US, EU and APAC, with static IP ranges
Self-hosting	No	Open-source runner on your own machines, free to use
Compliance and transparency	SOC 2 is in progress. We're open about where we stand: live status and data retention policy.	SOC 2

Every feature Battleships tracks (43)

Feature	bochka	GitHub Actions
Full VM (own kernel)	Yes	Yes
Docker inside	Yes	Yes
Nested virtualization	Yes	Yes
Root	Yes	Yes
Systemd	Yes	No
Desktop GUI	Yes	No
Browser	Yes	Yes
Computer-use API	Yes	No
SSH	Yes	No
Persistent disk	Yes	No
Pause/resume	Yes	No
Automatic snapshots	Yes	No
Fork/clone	Yes	No
Live fork (no pause)	Yes	No
Memory fork	No	No
Custom image (Docker or snapshot)	Yes	Yes
≥24 h sessions	Yes	No
Idle auto-stop	No	Yes
Wake on request	No	No
HTTPS preview URLs	Yes	No
Raw TCP inbound	Yes	No
Public IPv4	No	No
Egress allowlist	No	Yes
Secret proxy	Yes	No
Secret proxy for any API	Yes	No
Preinstalled agents	Yes	Yes
Agent harness API	Yes	Yes
MCP server	No	Yes
GPU	No	Yes
Windows	No	Yes
Arm64	No	Yes
Self-hosting	No	Yes
BYOC	No	No
Open source	No	Yes
Scoped API keys	Yes	Not published
Spend limits	Yes	Yes
Usage per sandbox (API)	Yes	Yes
Audit logs	No	Yes
SSO	No	Yes
SOC 2	No	Yes
HIPAA	No	No
EU data residency	Yes	Yes
Zero data retention	Yes	Not published

Performance

bochka26.69 runs/s

GitHub ActionsNot measured

There is no comparable measurement for GitHub Actions hosted runners in the hpc-sandbox-benchmarks CPU test. bochka did 26.69 runs per second on it at 4 vCPU / 8 GB, on dedicated AMD Ryzen 9 9950X hosts.

Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. GitHub Actions 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.

Scenario	bochka, a month	GitHub Actions, 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	$80 · Team · with limits
One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month	$26 · $20/mo	$354 · Team · 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	$3,814 · Team · 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	$1,092 · Enterprise Cloud · 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	$2,721 · Enterprise Cloud · with limits

bochka: Plans start at $20 a month, and the whole fee is usage credit. You pay per second while a sandbox runs, and a stopped sandbox costs nothing and keeps its disk.

GitHub Actions: GitHub bills per minute per job, rounding each job up to the next whole minute, at a rate set by the runner type, from $0.006 a minute for 2-core Linux. Plans include monthly minutes for standard runners only, and larger and GPU runners are always billed.

Limits in those scenarios

GitHub Actions does not fit any of the five scenarios as a sandbox: it is a CI service, every job ends after 6 hours, and nothing on the disk survives to the next job. For sandboxes for your users, its terms also rule out offering Actions as a service. For 1,000 short jobs at once, Enterprise Cloud stops at 500 concurrent jobs, and bochka's $2,000 plan queues starts above 90 a minute.

Scenario	bochka	GitHub Actions
One coding agent, workdays	None that bite	Team: sessions capped at 6 h missing persistent disk different kind of product: CI runner
One agent that never stops	None that bite	Team: sessions capped at 6 h missing persistent disk different kind of product: CI runner
50 agents in parallel	None that bite	Team: sessions capped at 6 h missing persistent disk different kind of product: CI runner
Sandboxes for your users	None that bite	different kind of product: CI runner
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)	Enterprise Cloud: max 500 concurrent different kind of product: CI runner

Which one should you use?

Use GitHub Actions when

you build, test and deploy a GitHub repository, especially for macOS, Windows or arm64 builds, or when you want the Copilot, Claude or Codex agents working on your issues and pull requests.

Native CI for GitHub repos. For building and testing a GitHub repository, Actions runs right from pull requests and pushes with no extra service to wire up.
macOS, Windows, arm64 and GPUs. Teams that build for Apple, Windows or arm64, or need a T4 GPU, get hosted runners for each.
Your own runners and your own network. You can run the open-source runner on your own machines for free, or join larger runners to your Azure network.

Use bochka when

an agent needs its own machine that keeps its disk, runs past 6 hours, takes inbound traffic, shows a desktop and keeps real API keys out of reach.

A machine that stays. On GitHub Actions every job starts on a clean VM and ends within 6 hours, so an agent reinstalls its tools and reclones the repo each time, and work past 6 hours has to be split across jobs.
Reach it from outside. Hosted runners accept no inbound connections, so a preview, a webhook or a live look at what the agent is doing needs a third-party tunnel inside the job.
Agents you drive from outside, keys kept out. In an Actions job, secrets arrive as environment variables inside the VM the agent controls, where one prompt injection can read them; hiding them from logs does not stop that.

Questions people ask

Can I use GitHub Actions as a sandbox for AI agents?

For agent work on a repository, yes: GitHub's own Copilot cloud agent runs on Actions runners. But each job is a fresh VM that ends within 6 hours, takes no inbound traffic, and GitHub's terms limit hosted runners to work on the repository's software project.

Is bochka cheaper than GitHub Actions?

In the five agent scenarios on this page bochka costs less, but GitHub Actions does not fit any of them, because it is a CI service with 6-hour jobs. For plain CI on a GitHub repository, its included minutes and free public-repo runners can make it the cheaper choice.

How long can a GitHub Actions job run?

Up to 6 hours per job on GitHub-hosted runners, and 15 minutes on the single-CPU slim runners. A bochka sandbox has no session cap on paid plans.

Do files persist between GitHub Actions runs?

Not on the runner. Each job starts on a new VM, and files carry over only through the cache, artifacts or a custom image. A bochka sandbox keeps its disk, saves it every minute and can be stopped and started again.

Can I run Docker and Android emulators in each?

Both run Docker, and both expose KVM on Linux for Android emulators. On GitHub runners you add the runner user to the KVM group first, and the job still ends within 6 hours.

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.

Start a sandbox

Install the CLI, then run bochka new.

Product preview