bochka vs Ellipsis
Ellipsis runs Claude Code or Codex sessions on your repos, wired into GitHub, Slack and Linear. bochka gives each agent a whole machine it can keep for days, reach over SSH and host apps from.
Where bochka stands out
No one-hour wall
A bochka sandbox has no session cap on paid plans, and its disk is saved every minute. An Ellipsis environment stops after one hour at most, because it never outlives its GitHub token.
A whole machine you can reach
Each bochka sandbox is a full VM with root, systemd, SSH, Docker and /dev/kvm. It can serve your app on HTTPS URLs for up to 50 ports or raw TCP, and it has a 60fps desktop with Chromium for computer use.
Keys stay out for every API
bochka's credentials proxy keeps real keys outside the sandbox for model providers and any HTTPS API. When one account runs out of credits, the credential can switch to another account of the same type.
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
- Regions
- EU: Germany, Finland, France
Ellipsis
Ellipsis is a cloud platform for coding agents: you build agents for feature work, background tasks and code review, and it runs them for you.
Ellipsis is not a general sandbox API. Each session is a conversation with Claude Code or Codex in its own cloud machine, started from the dashboard, a CLI, Python or TypeScript SDKs, a schedule, or a mention in GitHub, Slack or Linear. It handles permissions, logs and budgets for teams that want agents working on their repos, and its sandboxes run on Freestyle VMs.
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 | Ellipsis: 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 build, test and serve apps for hours; Ellipsis stops each environment after one hour and has no inbound ports, while bochka has neither limit. | 3rd cheapest of 367 $634 a month | 340th of 367, with limits $10,095 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 Ellipsis's one-hour environments; on bochka it simply keeps running. | 5th cheapest of 366 $657 a month | 342nd of 366, with limits $13,141 a month |
| Code interpreter bursts 100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month Short code runs fit easily inside Ellipsis's one-hour limit, while bochka's plans cap starts per minute, which hurts very bursty interpreter work. | 186th of 345, with limits $2,000 a month | 49th cheapest of 345 $1,198 a month |
| Sandboxes for your users 300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month Both can give each of your users a sandbox; bochka adds hosting, SSH and a desktop, while Ellipsis sleeps between messages and wakes on the next one. | Cheapest of 361 $209 a month | 29th cheapest of 361 $4,309 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 | 182nd of 360, with limits $20,816 a month |
| Agent evals 250 at once · 4 vCPU / 8 GB · 15,000 sessions of 20 min a month Both fit agent evals; bochka's machines run as long as an eval needs and can be forked mid-run to branch from the same state. | 10th cheapest of 364 $200 a month | 84th cheapest of 364 $3,349 a month |
| CI runners 40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month Ellipsis is built to review pull requests and fix code from GitHub, while bochka gives CI jobs a full VM with Docker and /dev/kvm for emulators. | 4th cheapest of 377 $86 a month | 127th cheapest of 377 $2,010 a month |
| 24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB A dev box needs SSH and has to stay up all day, and Ellipsis offers neither. | 18th cheapest of 368 $26 a month | 349th of 368, with limits $526 a month |
| Computer-use agents 50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month Computer-use agents need a desktop and a browser; bochka streams a 60fps desktop with Chromium, and Ellipsis runs headless coding agents only. | 6th cheapest of 369 $200 a month | 329th of 369, with limits $3,014 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 | 330th of 367, with limits $1,507 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 | 330th of 346, with limits $1,781 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 | 313th of 338, with limits $35,629 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 | 314th of 360, with limits $994 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 | Can't run it | Can't run it |
Key features
| Feature | bochka | Ellipsis |
|---|---|---|
| What it is | Sandbox API: a full VM per agent for any workload | Coding-agent platform: sessions run Claude Code or Codex on your repos |
| Isolation | Full VM with its own kernel on dedicated AMD Ryzen 9 9950X hosts | Its own isolated cloud machine per session, on Freestyle VMs |
| Biggest machine | Up to 16 vCPU / 32 GB on the $100 plan and above | Up to 32 vCPU / 64 GB |
| Session length | No cap on paid plans; you set the time to live | 60 seconds to 1 hour per environment; 1 hour is the maximum because it never outlives its GitHub token |
| Between tasks | Disk saved every minute and on stop; start it again by API | Goes cold between messages and restores the workspace, uncommitted changes included, on the next one |
| Fork | Fork a running sandbox with its disk | No fork or clone API |
| Coding agents | Seven agents preinstalled, with a prompt, steer and events API you drive from outside | Claude Code and Codex, with a sessions API, SDKs and CLI; switching harness is a one-line change |
| Triggers and schedules | Start agents from your own code through the API | Run agents on a schedule, from events, or from a mention in GitHub, Slack or Linear |
| Getting into the machine | SSH, root and systemd | Dashboard, CLI and API only; no SSH documented |
| Hosting | HTTPS URLs for up to 50 ports, token-protected by default; raw TCP | No inbound ports or URLs documented |
| Nested virtualization | Yes, /dev/kvm for Android emulators and VMs | Not offered |
| Desktop and browser | 60fps Linux desktop with Chromium and computer use | Headless; a browser only if your setup script installs one |
| Keeping API keys out of the sandbox | Credentials proxy for model providers and any HTTPS API; switch a credential to another account when credits run out | Model calls use a per-session key; other secrets are readable by the agent |
| Self-hosting | No | BYOC into your own AWS VPC |
| 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 | Ellipsis |
|---|---|---|
| Full VM (own kernel) | Yes | Yes |
| Docker inside | Yes | Yes |
| Nested virtualization | Yes | No |
| Root | Yes | Not published |
| Systemd | Yes | Not published |
| Desktop GUI | Yes | No |
| Browser | Yes | No |
| Computer-use API | Yes | No |
| SSH | Yes | No |
| Persistent disk | Yes | Yes |
| Pause/resume | Yes | Yes |
| Automatic snapshots | Yes | Yes |
| 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 | Yes |
| HTTPS preview URLs | Yes | No |
| Raw TCP inbound | Yes | Not published |
| Public IPv4 | No | No |
| Egress allowlist | No | No |
| Secret proxy | Yes | Yes |
| Secret proxy for any API | Yes | No |
| Preinstalled agents | Yes | Yes |
| Agent harness API | Yes | Yes |
| MCP server | No | No |
| GPU | No | No |
| Windows | No | No |
| Arm64 | No | No |
| Self-hosting | No | Yes |
| BYOC | No | Yes |
| Open source | No | No |
| Scoped API keys | Yes | Not published |
| Spend limits | Yes | Yes |
| Usage per sandbox (API) | Yes | No |
| Audit logs | No | Yes |
| SSO | No | No |
| SOC 2 | No | Yes |
| HIPAA | No | No |
| EU data residency | Yes | No |
| Zero data retention | Yes | Not published |
Performance
There is no comparable measurement for Ellipsis. bochka did 26.69 runs per second on the StarSling hpc-sandbox-benchmarks Node.js build-and-test loop at 4 vCPU / 8 GB, on dedicated AMD Ryzen 9 9950X hosts. Ellipsis has not been measured on the same test.
Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. Ellipsis 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 | Ellipsis, 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 | $133 · Managed usage · with limits |
| One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month | $26 · $20/mo | $550 · Managed usage · 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 | $6,631 · Managed usage · 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 | $1,884 · Managed usage |
| 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 | $4,163 · Managed usage |
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.
Ellipsis: Organization accounts pay for the CPU and memory allocated while a sandbox runs, including while it waits for your next message, from prepaid credits that carry a platform fee. Personal accounts get CPU and memory free for their own sessions, and model tokens are billed separately or come from your own Claude or ChatGPT subscription.
Limits in those scenarios
The limit that bites on Ellipsis is time: an environment stops after one hour, so a workday agent, an always-on agent and a 50-agent fleet all need restarts, which is why they do not fit. Ellipsis also publishes no price for disk or egress, so those are left out of its totals. bochka has no session cap, but at 1,000 short jobs at once its $2,000 plan queues starts, while Ellipsis has no published start limit.
| Scenario | bochka | Ellipsis |
|---|---|---|
| One coding agent, workdays | None that bite |
|
| One agent that never stops | None that bite |
|
| 50 agents in parallel | None that bite |
|
| Sandboxes for your users | None that bite |
|
| 1,000 short jobs at once |
|
|
Which one should you use?
Use Ellipsis when
you want ready-made coding agents that review pull requests and take tasks from GitHub, Slack or Linear, or that must run inside your own AWS account.
- Agents wired into your team's tools. Good for teams that want agents reviewing pull requests and taking tasks from GitHub, Slack or Linear without building that glue.
- Runs in your own AWS account. Its BYOC option deploys into your AWS VPC, so code and model calls stay inside your network.
- Sleeps between messages. A session goes cold when idle and comes back with its workspace on the next message, which suits chat-paced work.
Use bochka when
your agents need a real machine for as long as the work takes: SSH, Docker, emulators, a browser on a desktop, and apps served on their own URLs.
- No one-hour wall. A long build, a data job or an agent that works all afternoon gets cut off on Ellipsis and must be resumed in a new environment, losing anything that only lived in memory.
- A whole machine you can reach. On Ellipsis the agent can edit and test code, but you cannot SSH in, open a preview URL, run an emulator or let it use a browser on screen.
- Keys stay out for every API. Ellipsis keeps model keys out, but its docs say the agent can read every secret its environment sets, so a prompt injection can leak your other API keys.
Questions people ask
Is bochka cheaper than Ellipsis?
For organization accounts, bochka costs less in all five scenarios on this page: for example $26 against $550 a month for one agent that never stops. For 1,000 short jobs at once, bochka's plan still costs less but queues the starts, while Ellipsis has no published start limit. Personal Ellipsis accounts pay nothing for CPU and memory on their own sessions.
Can an Ellipsis sandbox run for more than an hour?
No. An Ellipsis environment's timeout is 60 seconds to 1 hour, and 1 hour is the maximum because it never outlives its GitHub token. A follow-up message continues in a new environment with the same workspace. bochka has no session cap on paid plans.
Is Ellipsis a sandbox provider?
Not in the usual sense. Ellipsis is a platform for coding agents that runs each Claude Code or Codex session in its own cloud machine. bochka is a sandbox API: you get the machine itself and can run any agent or service on it.
Can I run Claude Code in the cloud with bochka, like on Ellipsis?
Yes. bochka comes with Claude Code, Codex and five other agents preinstalled, and you prompt, steer and follow them from outside through one API. Ellipsis adds ready-made GitHub, Slack and Linear triggers, which you would wire up yourself on bochka.
Can I SSH into an Ellipsis sandbox or open a preview URL?
Ellipsis documents no SSH access and no inbound ports; you reach sessions through its dashboard, CLI and API. bochka gives every sandbox SSH and HTTPS URLs for up to 50 ports.
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