Beam vs Google Agent Runtime
Beam is a serverless GPU cloud with quick gVisor sandboxes you can also self-host. Google Agent Runtime, once called Vertex AI Agent Engine, hosts your agent code inside Google Cloud and gives it locked sandboxes that run without root.
The companies and products
Beam
Beam is a serverless cloud for GPU inference, sandboxes and background jobs, built on its open-source runtime Beta9.
Beam runs functions, REST endpoints, task queues, pods, sandboxes and VMs on CPUs and GPUs, all from a Python or TypeScript SDK. A sandbox is a container isolated with gVisor that starts in a second or two, lives until you terminate it or its time to live runs out, and can be saved as a filesystem image or a memory snapshot. Beam VMs are a separate CPU product: a microVM with its own kernel, systemd and a root disk that stays when it stops. Much of Beam is a GPU serverless platform for AI teams.
- Bring your own cloud
- Management fee per vCPU and GB of RAM, on top of your own cloud bill
Google Agent Runtime
Google Agent Runtime is the agent hosting part of Gemini Enterprise Agent Platform on Google Cloud, renamed from Vertex AI Agent Engine in April 2026.
It is a hyperscaler service, not a sandbox company. Agent Runtime deploys agents you build with ADK, LangGraph, LangChain, AG2 or LlamaIndex and bills them only while they work on a turn. Next to it, Agent Sandbox gives those agents a place to run code: a Code Execution sandbox, a shell or custom container sandbox, and a Computer Use browser. All of them run as secure containers inside your Google Cloud project.
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. “8th cheapest of 345” means 7 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 workload | Beam: price rank | Google Agent Runtime: price rank | boat: price rank |
|---|---|---|---|
| Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month A software factory needs Docker, systemd and /dev/kvm: Beam sandboxes have Docker but no systemd or /dev/kvm, Google's have none of them, and boat has all three on every VM. | 69th of 367, with limits $551 a month | 295th of 367, with limits $5,801 a month | 3rd cheapest of 367 $634 a month |
| Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB Google fits a long-lived assistant better than a Beam sandbox because Agent Runtime does not bill idle time between turns, though its sandboxes are deleted when their TTL ends; a boat VM has no session cap on paid plans. | 115th of 366, with limits $651 a month | 28th cheapest of 366 $7,519 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 Beam fits short code runs well: its sandboxes start in a second or two and bill per second, while Google's Code Execution sandbox is a ready Python interpreter with no network. | 8th cheapest of 345 $117 a month | 254th of 345, with limits $275 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 Beam gives each user a quick container with a public HTTPS URL per port, while a Google sandbox has no public URL; boat gives each user a full VM with token-protected HTTPS on up to 50 ports. | 3rd cheapest of 361 $290 a month | 280th of 361, with limits $1,916 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 | 341st of 360, with limits $939 a month | 128th of 360, with limits $8,240 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 | 13th cheapest of 364 $251 a month | 79th cheapest of 364 $2,060 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 Beam can run Docker builds with docker_enabled and Google's unprivileged sandboxes cannot; jobs that need /dev/kvm, like Android builds, run only on boat. | 12th cheapest of 377 $171 a month | 98th cheapest of 377 $989 a month | 4th cheapest of 377 $86 a month |
| 24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB A Beam VM keeps its 50 GiB root disk while stopped, while Google has no SSH and runs each command in a new shell; boat gives you SSH, a desktop and a disk saved every minute. | 13th cheapest of 368 $21 a month | 311th of 368, with limits $241 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 Google gives agents a browser with a VNC view and a click and type API, Beam has no browser product, and boat streams a full 60fps Linux desktop with Chromium. | 111th of 369, with limits $237 a month | 283rd of 369, with limits $1,648 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 | 195th of 367, with limits $176 a month | 102nd of 367, with limits $824 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 | 200th of 346, with limits $131 a month | 53rd cheapest of 346 $515 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 | 42nd of 338, with limits $923 a month | 210th of 338, with limits $8,240 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 | 175th of 360, with limits $155 a month | 260th of 360, with limits $412 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 Only Beam has GPUs here, from RTX 4090 to H100 and B200; Google's sandboxes and boat are CPU only. | Cheapest of 98 $458 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 | Can't run it |
Key features
| Feature | Beam | Google Agent Runtime | boat |
|---|---|---|---|
| Isolation | gVisor container by default; separate Beam VMs are CPU microVMs with their own kernel | Secure container sandboxing inside your Google Cloud project | Every sandbox is a full VM with its own kernel on dedicated AMD Ryzen 9 9950X hosts |
| Root, Docker and /dev/kvm | Root prompt and Docker with docker_enabled; systemd only in Beam VMs; no /dev/kvm documented | Runs as appuser with no sudo; images must not need root; no Docker or /dev/kvm documented | Root, systemd, Docker and /dev/kvm on every machine |
| GPUs | GPU sandboxes and serverless GPUs up to H100, H200 and B200 | No GPUs in sandboxes | No, CPU only |
| How long a sandbox lives | Default 10 minutes; set -1 to run until you terminate it | TTL of up to 7 days for a custom container and 14 days for Code Execution | No session cap on paid plans; you set the time to live |
| Files after a stop | A sandbox ends with its time to live unless you save an image or use a volume; a Beam VM keeps a 50 GiB root disk while stopped | Pause keeps the file system at a lower rate Google does not publish; the sandbox is deleted when its TTL ends | Disk saved every minute and on stop; a stopped sandbox keeps it for free |
| Memory snapshots and branching | Memory snapshot of a running sandbox with its processes, restored into new sandboxes | Snapshots keep memory and files; start many sandboxes from one snapshot | Fork a running sandbox with its disk; resume reboots from disk (no memory snapshots) |
| Desktop and browser | No desktop or browser product | A browser in a container with a VNC live view, a click, type and screenshot API, and CDP | Linux desktop streamed at 60fps with Chromium, computer use and browser control |
| Reaching the sandbox | A public HTTPS URL per exposed port with no built-in auth; no raw TCP or dedicated IP | Through Google's API or a private endpoint in your own VPC; no public URL documented | HTTPS URLs for up to 50 ports, token-protected by default; raw TCP; dedicated IPv4 or IPv6 |
| Outbound rules | Block all outbound traffic, or allow up to 10 CIDR ranges, changeable while running | Off unless the template turns it on; can route through your VPC with Cloud NAT, a static IP, Cloud NGFW or Secure Web Proxy | Open internet on every plan; no provider-side allowlist |
| Coding agents | None preinstalled; you install and run them through process exec | Managed Agents API (Preview) runs Google's Antigravity harness | Claude Code, Codex and five other agents preinstalled, with a prompt, steer and events API |
| Keeping API keys out of the sandbox | Secrets are exposed as environment variables inside the sandbox | You provision and scope the credentials the agent uses | Credentials proxy for model providers and any HTTPS API; switch a credential to another account when credits run out |
| Hosting your agent code | Serverless functions, REST endpoints and task queues from a Python or TypeScript SDK | Agent Runtime deploys ADK, LangGraph, LangChain, AG2 and LlamaIndex agents; idle time between turns is not billed | You run your own agent loop inside a sandbox |
| Regions | US only; no other region is listed | 23 locations in the US, Europe, Asia, Canada, Australia and Brazil | EU only: Germany, Finland, France |
| Self-hosting | Beta9 runtime is AGPL-3.0 and free to self-host; bring your own cloud for a fee | No; it runs only as a Google Cloud service | No |
Every feature Battleships tracks (43)
| Feature | Beam | Google Agent Runtime | boat |
|---|---|---|---|
| Full VM (own kernel) | No | No | Yes |
| Docker inside | Yes | No | Yes |
| Nested virtualization | No | No | Yes |
| Root | Yes | No | Yes |
| Systemd | No | Not published | Yes |
| Desktop GUI | No | No | Yes |
| Browser | No | Yes | Yes |
| Computer-use API | No | Yes | Yes |
| SSH | Yes | No | Yes |
| Persistent disk | Yes | Yes | Yes |
| Pause/resume | No | Yes | Yes |
| Automatic snapshots | No | Yes | Yes |
| Fork/clone | Yes | Yes | Yes |
| Live fork (no pause) | Yes | Yes | Yes |
| Memory fork | Yes | Yes | No |
| Custom image (Docker or snapshot) | Yes | Yes | Yes |
| ≥24 h sessions | Yes | Yes | Yes |
| Idle auto-stop | No | No | No |
| Wake on request | No | No | No |
| HTTPS preview URLs | Yes | No | Yes |
| Raw TCP inbound | No | Not published | Yes |
| Public IPv4 | No | No | No |
| Egress allowlist | Yes | Yes | No |
| Secret proxy | No | No | Yes |
| Secret proxy for any API | No | No | Yes |
| Preinstalled agents | No | Yes | Yes |
| Agent harness API | No | Yes | Yes |
| MCP server | Yes | No | No |
| GPU | Yes | No | No |
| Windows | No | No | No |
| Arm64 | No | Yes | No |
| Self-hosting | Yes | No | No |
| BYOC | Yes | No | No |
| Open source | Yes | No | No |
| Scoped API keys | Not published | Not published | Yes |
| Spend limits | Not published | Not published | Yes |
| Usage per sandbox (API) | No | Yes | Yes |
| Audit logs | Not published | Not published | No |
| SSO | No | Yes | No |
| SOC 2 | Yes | Yes | No |
| HIPAA | Yes | Yes | No |
| EU data residency | Not published | Yes | Yes |
| Zero data retention | Not published | Not published | Yes |
Performance
There is no comparable measurement for Beam or Google Agent Runtime: neither has a result on the StarSling hpc-sandbox-benchmarks CPU test, so we cannot say which is faster at CPU work. boat did 26.69 runs per second on that test at 4 vCPU / 8 GB. Beam's own guide says its sandboxes cold-boot in 1 to 3 seconds.
Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. Beam and Google Agent Runtime 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 | Beam, a month | Google Agent Runtime, a month | boat, 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 | $5 · Developer | $73 · Pay as you go | $20 · $20/mo |
| One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month | $22 · Developer | $301 · Pay as you go | $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 | $410 · Team | $3,626 · Pay as you go · 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 | $191 · Team | $1,030 · Pay as you go | $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 | $259 · Team | $2,060 · Pay as you go | $2,000 · $2000/mo · with limits |
Beam: Beam sandboxes bill per second for each CPU core (two vCPU) and GiB of RAM you request, until you terminate them or their time to live ends, so an idle tail is billed too. Developer has no fee, and Team costs $89 a month plus usage, with no egress fees.
Google Agent Runtime: Google bills $0.085 per vCPU-hour and $0.009 per GiB-hour allocated, with the first 50 vCPU-hours and 100 GiB-hours free each month. Runtime skips idle time between turns, but sandboxes bill while they exist, and disk and egress rates are not published.
Limits in those scenarios
Beam's Developer plan allows 30 CPU containers at once, so 50 parallel agents need the Team plan. Google cannot run the 50-agent fleet because its sandboxes have no Docker inside, and a Google sandbox is deleted when its TTL ends, 7 days at most for a custom container. Neither publishes a disk price that fits these scenarios, so no disk cost is counted for either. boat fits every scenario except the 1,000-job burst, where its start limits queue the jobs.
| Scenario | Beam | Google Agent Runtime | boat |
|---|---|---|---|
| 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 Beam when
your agents need GPUs, when you want quick per-second sandboxes with Docker for short jobs, when you want low-cost CPU VMs in the US, or when you must self-host.
- GPUs in the same platform. For agents or jobs that need a GPU, Beam offers GPU sandboxes and serverless GPUs from RTX 4090 to H100 and B200.
- Quick sandboxes with Docker inside. For short code runs and builds, Beam sandboxes start in a second or two, bill per second and can run Docker.
- Open source and your own cloud. For teams that must self-host, the Beta9 runtime is AGPL-3.0 and free to run, or Beam can run in your cloud.
Use Google Agent Runtime when
you build agents with ADK or LangGraph on Google Cloud and want them hosted inside your own project, network rules and regions, with short code runs in a locked sandbox.
- A managed home for your agent code. For agents built with ADK or LangGraph, Agent Runtime hosts them and does not bill idle time between turns.
- Inside your Google Cloud network. For agents that must stay in a Google Cloud perimeter, egress is off by default and can route through your own firewall.
- Regions worldwide and memory snapshots. For work outside the US, Google runs in 23 locations, and a snapshot can branch a sandbox with its memory.
Where boat fits
Beam is strongest when you need GPUs or quick throwaway containers, and Google is strongest when your agent code should live inside Google Cloud. If what you are really running is coding agents that install tools, run Docker, serve what they build and keep working for days, boat gives each one a full VM with all of that built in.
A full VM with root, Docker and /dev/kvm
A Beam sandbox is a gVisor container with no systemd or /dev/kvm, and a Google sandbox runs as appuser with no sudo or Docker. On boat, apt installs, docker compose and Android emulators just work.
Public URLs with access control
Beam gives each port a public URL anyone can open, with no raw TCP, and Google has no public URL at all, so you set up a VPC endpoint or tunnel. boat serves token-protected HTTPS on 50 ports, plus raw TCP.
Agents, desktop and keys built in
Beam has no agents or desktop and puts secrets in environment variables; Google runs only its Antigravity harness and a browser. boat ships Claude Code and Codex, a 60fps desktop, and a credentials proxy that keeps keys outside.
Use boat when your agents need a real computer: root, Docker, SSH, a desktop and public URLs on a full VM that keeps its disk, with Claude Code, Codex and other coding agents built in. Compare it one to one: boat vs Beam and boat vs Google Agent Runtime.
Questions people ask
Is Beam cheaper than Google Agent Runtime?
Yes, in all five scenarios we priced. One coding agent on workdays comes to $5 a month on Beam and $73 on Google, one agent that never stops $22 against $301, and sandboxes for your users $191 against $1,030. Google's numbers assume a sandbox that exists all those hours; a Runtime agent that sits idle between turns bills less.
Which is faster, Beam or Google Agent Runtime?
We cannot say: neither has a result on the StarSling CPU benchmark we use. Beam's guide says its sandboxes cold-boot in 1 to 3 seconds. boat did 26.69 runs per second on that benchmark at 4 vCPU / 8 GB.
Can Beam or Google run Docker inside a sandbox?
Beam can, with docker_enabled on a sandbox. Google cannot: shell commands run as the unprivileged user appuser with no sudo, and custom images must not need root. boat runs Docker, systemd and /dev/kvm on every VM.
Is Google Agent Runtime the same as Vertex AI Agent Engine?
Yes. Google renamed Vertex AI Agent Engine to Agent Runtime in April 2026, as part of Gemini Enterprise Agent Platform. Its code, shell and browser sandboxes are a separate part called Agent Sandbox.
Which one has GPUs?
Only Beam. It offers GPU sandboxes and serverless GPUs from T4 and RTX 4090 up to H100, H200 and B200. Google's agent sandboxes and boat are CPU only.
Where does my data run?
Beam runs in the United States on AWS and Google Cloud, with no other region listed. Google runs in 23 locations across the US, Europe, Asia, Canada, Australia and Brazil. boat runs only in the EU, in Germany, Finland and France.
Try boat
Install the CLI, then run boat new.
7-day free trial