Agent 37 vs Islo
Agent 37 gives each user a light gVisor sandbox that sleeps when idle and keeps its memory, with a full VM type as an extra. Islo gives every coding agent a microVM with its own kernel and a gateway that decides what it can reach and logs each call.
The companies and products
Agent 37
Agent 37 hosts AI agents, one isolated, always-on sandbox per user, behind one API.
You create an instance from a template (Hermes, OpenClaw, Claude Code, Codex, OpenCode, Pi, Grok or your own Docker image) and talk to it through a chat API, files, exec, SSH and HTTPS URLs. The default instance is a container enforced by gVisor; a VM type gives a full Linux machine with its own kernel for Docker, Kubernetes and Android emulators. It is built for companies that give each of their customers an agent.
- Where it runs
- Dedicated servers in the United States
- Instance types
- Default (shared), Performance (dedicated cores), VM
- Compliance
- SOC 2 Type I, Type II in progress
Islo
Islo is an agent sandbox from Incredibuild: each coding agent gets its own microVM, with network rules and credentials handled at a gateway.
An Islo sandbox is a Linux microVM with its own kernel, built from a container image, with Claude Code, Cursor CLI and Codex preinstalled. Outbound traffic can pass through a gateway that allows or blocks hosts, injects credentials and logs every request. It is built for teams that run coding agents, evals and RL rollouts in the background, on Islo's cloud or in their own cloud account.
- Launched
- May 4, 2026
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. “2nd cheapest of 367” means 1 of the other 366 providers runs 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 | Agent 37: price rank | Islo: price rank | bochka: price rank |
|---|---|---|---|
| Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month Islo runs Docker in every microVM, while Agent 37 needs its VM type for Docker at 2.5 times the compute rate; bochka adds systemd, /dev/kvm and a desktop on every VM. | 2nd cheapest of 367 $551 a month | 326th of 367, with limits $7,402 a month | 3rd cheapest of 367 $634 a month |
| Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB Both can sleep an idle assistant with its memory and wake it on the next request; Agent 37 is built for one always-on assistant per user and costs far less to keep. | Cheapest of 366 $279 a month | 30th cheapest of 366 $9,928 a month | 5th cheapest of 366 $657 a month |
| Code interpreter bursts 100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month Agent 37's small default sandboxes suit short code runs; Islo needs a raised concurrency limit for big bursts, and bochka's start limits queue very large ones. | Cheapest of 345 $22 a month | 153rd of 345, with limits $503 a month | 186th 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 Agent 37 stops at 200 instances, counting sleeping and stopped ones, and Islo's Team plan at 50 running, so 300 at once fits neither; bochka plans go up to 2,000. | 30th of 361, with limits $79 a month | 171st of 361, with limits $3,160 a month | Cheapest of 361 $209 a month |
| RL rollouts 1,000 at once · 2 vCPU / 4 GB · 1,000,000 sessions of 3 min a month | 6th of 360, with limits $363 a month | 167th of 360, with limits $15,265 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 Islo has a built-in Harbor integration and gateway rules that stop trials from looking up answers; bochka runs evals on more sandboxes at once on full VMs. | 90th of 364, with limits $58 a month | 198th of 364, with limits $2,456 a month | 10th cheapest of 364 $200 a month |
| CI runners 40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month | 2nd cheapest of 377 $84 a month | 119th cheapest of 377 $1,507 a month | 4th cheapest of 377 $86 a month |
| 24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB All three keep a machine and its disk around the clock; Agent 37 is the cheapest to keep idle, and bochka's full VM adds systemd and raw TCP. | 3rd cheapest of 368 $12 a month | 112th cheapest of 368 $402 a month | 18th cheapest of 368 $26 a month |
| Computer-use agents 50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month Agent 37 adds a noVNC desktop through a template build, Islo has no desktop or computer-use API, and bochka streams a 60fps desktop on every sandbox. | Cheapest of 369 $53 a month | 310th of 369, with limits $2,210 a month | 6th cheapest of 369 $200 a month |
| Browser-use agents 50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month | 48th of 367, with limits $29 a month | 113th of 367, with limits $1,105 a month | 7th 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 | Cheapest of 346 $32 a month | 66th cheapest of 346 $748 a month | 6th 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 | 36th of 338, with limits $648 a month | 265th of 338, with limits $14,966 a month | 71st 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 Agent 37 has no fork and Islo starts new sandboxes from a file snapshot; bochka can fork a running VM to try many paths at once. | Cheapest of 360 $43 a month | 47th of 360, with limits $746 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 it | 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 | Can't run it |
Key features
| Feature | Agent 37 | Islo | bochka |
|---|---|---|---|
| Isolation | gVisor container by default; full VM with its own kernel as an opt-in type | Every sandbox is a microVM with its own kernel | Every sandbox is a full VM with its own kernel |
| Docker inside | Only on the VM type, at 2.5x the default compute rate | Docker inside every microVM, with sudo | Docker, root and systemd in every sandbox |
| Nested virtualization | /dev/kvm on the VM type, with a ready Android template | No /dev/kvm or nested virtualization documented | /dev/kvm in every sandbox, Android emulators work |
| Sleep with memory | Opt-in auto-sleep after idle time, memory checkpointed, wakes on the next request | Pause keeps files, processes and memory; can pause after idle time and resume on the next exec, file access or webhook | Disk kept on stop and a stopped sandbox costs nothing, but resume reboots from disk |
| Fork and snapshots | No fork; daily backups plus one on-demand backup, restored in place only | Snapshot a running sandbox's files, then start new sandboxes from it | Fork a running sandbox; snapshots every minute; named snapshots as shareable templates |
| Custom images | Your own linux/amd64 Docker image as a template | Sandboxes built from a container image | Templates from named snapshots, no Dockerfile builds |
| Desktop and computer use | noVNC desktop through a template build or the Ubuntu VM template | A browser inside for testing; no desktop or computer-use API | 60fps desktop and Chromium, computer-use API, embeddable stream |
| Outbound internet rules | Open internet; no egress allowlist | Gateway allows or denies by host, path and HTTP method, with default deny and rate limits | Open internet; no provider-side allowlist |
| Keys kept out of the sandbox | Managed models, search and Composio through a per-instance token; your own keys go in as env vars | Gateway injects stored credentials into outbound requests for any host | Proxy for model providers and any HTTPS API you register; switch a credential to another account when credits run out |
| Reaching the sandbox | HTTPS URL per port, signed URLs, custom domains; no public IP; SSH through a WebSocket tunnel | HTTPS share links open to anyone with the URL, expiring after a TTL (24 hours by default); SSH through the CLI | HTTPS URLs for up to 50 ports, token-protected; raw TCP; its own IP |
| Built-in agents | Hermes, OpenClaw, Claude Code, Codex, Grok, OpenCode, Pi, with a chat API per instance | Claude Code, Cursor CLI and Codex; factory lines run on a schedule, webhook or GitHub, Slack and Linear events | Claude Code, Codex, pi, OpenCode, Prime Agent, Kimi Code, Mistral Vibe, with a prompt, steer and events API |
| Request logs | No per-request log of outbound calls documented | Per-request gateway logs of what each agent called and what was blocked, streamed to your SIEM | Usage API and spend limits; no per-request log of outbound calls |
| Sandboxes at once | 1 to 200 instances by lifetime top-ups, counting sleeping and stopped ones | Free 5 and Team 50 running sandboxes; paused ones do not count; more by contract | 100 to 2,000 running sandboxes by plan |
| Where it runs | Dedicated servers in the United States | US West, or inside your own cloud account with BYOC | EU only: Germany, Finland, France, with EU data residency |
Every feature Battleships tracks (43)
| Feature | Agent 37 | Islo | bochka |
|---|---|---|---|
| Full VM (own kernel) | No | Yes | Yes |
| Docker inside | No | Yes | Yes |
| Nested virtualization | No | No | Yes |
| Root | Yes | Yes | Yes |
| Systemd | Not published | Not published | Yes |
| Desktop GUI | Yes | No | Yes |
| Browser | Yes | Yes | Yes |
| Computer-use API | Yes | No | Yes |
| SSH | Yes | Yes | Yes |
| Persistent disk | Yes | Yes | Yes |
| Pause/resume | Yes | Yes | Yes |
| Automatic snapshots | Yes | Yes | Yes |
| Fork/clone | Yes | Yes | Yes |
| Live fork (no pause) | Yes | Yes | Yes |
| Memory fork | No | No | No |
| Custom image (Docker or snapshot) | Yes | Yes | Yes |
| ≥24 h sessions | Yes | Yes | Yes |
| Idle auto-stop | Yes | Yes | No |
| Wake on request | Yes | Yes | No |
| HTTPS preview URLs | Yes | Yes | Yes |
| Raw TCP inbound | Not published | No | Yes |
| Public IPv4 | No | No | No |
| Egress allowlist | No | Yes | No |
| Secret proxy | Yes | Yes | Yes |
| Secret proxy for any API | No | Yes | Yes |
| Preinstalled agents | Yes | Yes | Yes |
| Agent harness API | Yes | Yes | Yes |
| MCP server | No | No | No |
| GPU | No | No | No |
| Windows | No | No | No |
| Arm64 | No | No | No |
| Self-hosting | No | No | No |
| BYOC | No | Yes | No |
| Open source | No | No | No |
| Scoped API keys | Not published | Not published | Yes |
| Spend limits | Yes | Not published | Yes |
| Usage per sandbox (API) | Yes | Yes | Yes |
| Audit logs | Not published | Yes | No |
| SSO | No | Yes | No |
| SOC 2 | Yes | Not published | No |
| HIPAA | No | Not published | No |
| EU data residency | No | No | Yes |
| Zero data retention | Not published | Not published | Yes |
Performance
There is no comparable measurement for Agent 37 or Islo: neither was in StarSling's public hpc-sandbox-benchmarks run. In that run of 2026-09-29, bochka did 26.69 Node.js build-and-test runs per second at 4 vCPU / 8 GB on dedicated AMD Ryzen 9 9950X hosts, the fastest result there. Agent 37 says its default instances share host resources and sells a Performance type on dedicated cores.
Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. Agent 37 and Islo have 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 | Agent 37, a month | Islo, a month | bochka, 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 | $6 · Wallet (any top-up) | $101 · Free | $20 · $20/mo |
| One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month | $12 · Wallet (any top-up) | $420 · Free | $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 | $329 · Wallet ($100 lifetime top-ups) | $4,863 · Team | $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 | $33 · Wallet ($250 lifetime top-ups) · with limits | $1,382 · Team · with limits | $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 | $73 · Wallet ($250 lifetime top-ups) · with limits | $3,053 · Team · with limits | $2,000 · $2000/mo · with limits |
Agent 37: A prepaid wallet with no plan fee, metered per minute: $0.80 per vCPU and $0.70 per GB of RAM per month while running, and $0.09 per GB of disk per month even while stopped or asleep. VM compute costs 2.5x and Performance compute 4x the default rate.
Islo: Islo bills each CPU core and each GB of memory and disk while a sandbox runs, with no plan fee and signup credits with no card. An idle sandbox keeps billing until a pause-after-idle timer fires, and that timer can only be set at creation.
Islo: A monthly plan from $20, all of it usage credit spent per second on running sandboxes. A stopped sandbox costs nothing and keeps its disk.
Limits in those scenarios
Agent 37 caps instances at 200 on its top self-serve tier, counting sleeping and stopped ones, and Islo's Team plan is reported at 50 running sandboxes, so neither fits 300 sandboxes for your users or 1,000 short jobs at once without a custom deal. Agent 37 is priced on its default gVisor type except for 50 agents in parallel, which needs Docker and so uses the VM type at 2.5x compute. Neither publishes an egress price, so their totals leave egress out. bochka fits 300 at once, but its start limits stop the 1,000-job burst.
| Scenario | Agent 37 | Islo | bochka |
|---|---|---|---|
| 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 Agent 37 when
you host one mostly idle assistant per customer in the US that should sleep with its memory kept, and you want the lowest bill.
- Built for one assistant per user. Useful for hosting a mostly idle assistant per customer: it sleeps, keeps its memory and wakes on the next message.
- Low cost to keep machines around. Agent 37 bills per minute from a prepaid wallet, with no plan fee, and comes out lowest in most scenarios below.
- Your Dockerfile, dedicated cores. Agent 37 builds templates from your own Docker image and sells a Performance type on dedicated cores.
Use Islo when
your coding agents need a microVM with Docker, per-host egress rules with a log of every call, or sandboxes inside your own cloud account.
- Egress rules and request logs. Useful if you must control what agents reach: Islo allows or blocks each host, path and method, and logs every call.
- A VM with Docker by default. Every Islo sandbox is a microVM with its own kernel and Docker inside, with no separate VM type to pay extra for.
- Runs in your own cloud. Useful if sandboxes must stay in your own cloud account: Islo can deploy its control plane and sandboxes there with BYOC.
Where bochka fits
If your agents do hours of real work, like installs, builds, Docker, emulators and UI checks, look at bochka. Every sandbox is a full VM with systemd, /dev/kvm and a 60fps desktop, you can fork one while it runs, and plans go up to 2,000 sandboxes at once in the EU.
Docker, systemd and /dev/kvm on every machine
On Agent 37 Docker needs the VM type at 2.5 times the compute rate, with no init system. Islo runs Docker but documents no /dev/kvm, so Android emulators and tools that start their own VMs stop at boot.
A desktop agents can see and use
Islo sandboxes are headless with no computer-use API, and Agent 37 needs a noVNC template build for a screen. bochka streams a 60fps desktop with Chromium and a computer-use API in every sandbox.
Fork a running machine, host on stable ports
Agent 37 has no fork and Islo copies only files, so each new try rebuilds running state. bochka forks a live VM and serves up to 50 token-protected HTTPS ports plus raw TCP, where Islo share links expire.
Use bochka when your agents need a full VM with systemd, /dev/kvm and a 60fps desktop, forks of running machines, ports you can host on, and up to 2,000 sandboxes at once in the EU. Compare it one to one: bochka vs Agent 37 and bochka vs Islo.
Questions people ask
Is Agent 37 cheaper than Islo?
Yes, in every scenario we priced. One coding agent on workdays comes to $6 a month on Agent 37 and $101 on Islo, one agent running all month to $12 versus $420, and 50 agents in parallel to $329 versus $4,863. bochka lands at $20, $26 and $317 for the same three.
Which is faster, Agent 37 or Islo?
There is no public benchmark for either, so we can't say. Agent 37 sells a Performance type on dedicated cores for steady speed. bochka did 26.69 runs per second on StarSling's Node.js build-and-test run at 4 vCPU / 8 GB.
Can Agent 37 and Islo run Docker?
Islo runs Docker inside every microVM. Agent 37 runs Docker only on its VM type, which bills compute at 2.5x; the default gVisor type cannot run a Docker daemon. bochka runs Docker in every sandbox.
Are they containers or VMs?
An Agent 37 sandbox is a gVisor container by default, with a full VM as an opt-in type. Every Islo sandbox is a microVM with its own kernel. Every bochka sandbox is a full VM with systemd and /dev/kvm.
Can an idle agent sleep and keep its memory?
Yes on both. Agent 37 checkpoints memory on auto-sleep and wakes on the next request, and Islo pauses with files, processes and memory and resumes on the next exec, file access or webhook. bochka keeps the disk when stopped and costs nothing, but reboots on resume.
Where do Agent 37 and Islo run?
Agent 37 runs on dedicated servers in the United States. Islo runs in US West, or inside your own cloud account with BYOC. bochka runs only in the EU, in Germany, Finland and France.
Try bochka
Install the CLI, then run bochka new.
Product preview