Jarvislabs vs Baseten
Jarvislabs rents you whole machines: GPU VMs, GPU containers and CPU VMs with root, billed by the minute, mostly in India. Baseten runs your models for you as containers that scale with traffic and down to zero, in the US and the EU. Baseten also launched Sandboxes for agents on November 5, 2026, free until November 16.
The companies and products
Jarvislabs
Jarvislabs is an AI cloud from India that rents Nvidia GPUs by the minute for training and serving models.
You can rent a GPU VM with root and your own kernel, a Template (a GPU container with PyTorch, ComfyUI or similar ready, opened in JupyterLab, VS Code or SSH), a plain CPU VM, or a multi-node GPU cluster. Shared Filesystems attach to several instances and outlive any one of them. It also serves open models through Managed Endpoints, Serverless workers and a token-priced Model API. It is built for people who train, fine-tune and serve AI models. Its jl CLI and skill file let a coding agent on your own machine drive the GPUs.
Baseten
Baseten is an inference cloud: it runs AI models on GPUs for production apps, sells hosted open models and training, and now runs agent sandboxes.
You package a model with Truss, its open-source format, or bring a Docker image, and Baseten runs it as replicas on Kubernetes that scale up with traffic and down to zero. It also sells Model APIs (open models priced per token) and training jobs on the same GPUs. Sandboxes, isolated environments where an agent runs commands and works with files, launched on November 5, 2026. They are free until November 16, 2026, and later pricing is not published. This page compares boat with both the CPU instances and the Sandboxes.
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. “18th cheapest of 374” means 17 of the other 373 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 | Jarvislabs: price rank | Baseten: price rank | boat: price rank |
|---|---|---|---|
| Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month A coding agent needs root, Docker and a kept disk. Jarvislabs VMs have root and a kept disk. Baseten Sandboxes keep files on standby and in volumes but do not document root or Docker. boat adds the agents themselves, preinstalled. | 185th of 365, with limits $1,406 a month | 234th of 365, with limits $2,619 a month | 5th cheapest of 365 $634 a month |
| Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB | 234th of 361, with limits $1,670 a month | 278th of 361, with limits $3,258 a month | 6th cheapest of 361 $657 a month |
| Code interpreter bursts 100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month Both bill per minute with no plan fee, but a Jarvislabs VM takes about 90 seconds to boot, and a Baseten replica pays for its cold start and, by default, 15 idle minutes. | 107th of 336, with limits $169 a month | 313th of 336, with limits $1,856 a month | 189th of 336, 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 Jarvislabs does not publish how many machines can run at once, and Baseten Sandboxes run 10 per organization at once by default, with more through support. So 300 sandboxes for your users means asking either one first. | 210th of 349, with limits $539 a month | 273rd of 349, with limits $1,868 a month | 3rd cheapest of 349 $209 a month |
| RL rollouts 1,000 at once · 2 vCPU / 4 GB · 1,000,000 sessions of 3 min a month Jarvislabs CPU VMs bill by the minute with root, which suits RL rollouts better than Baseten's model containers. Jarvislabs ranks 34th, boat 29th (start caps are its limit) and Baseten 190th; none fully fits. | 47th of 349, with limits $2,568 a month | 315th of 349, with limits $31,140 a month | 39th of 349, with limits $2,000 a month |
| Agent evals 250 at once · 4 vCPU / 8 GB · 15,000 sessions of 20 min a month | 118th of 359, with limits $411 a month | 297th of 359, with limits $1,609 a month | 10th cheapest of 359 $200 a month |
| CI runners 40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month Jarvislabs gives root on a CPU VM, so a CI runner with Docker works there. Baseten model containers run as a non-root user, and its sandbox docs do not document root or Docker. boat gives root and Docker on every VM and ranks 2nd, Jarvislabs 13th. | 18th cheapest of 374 $253 a month | 323rd of 374, with limits $1,119 a month | 2nd cheapest of 374 $86 a month |
| 24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB A Jarvislabs CPU VM keeps its disk as a 24/7 dev box ($68 a month). Baseten CPU replicas start fresh from their image; its Sandboxes keep files on standby and in volumes, but have no published price after November 16, 2026. boat comes to $26 and keeps a stopped disk for at least 14 days. | 55th cheapest of 371 $68 a month | 186th of 371, with limits $130 a month | 19th cheapest of 371 $26 a month |
| Computer-use agents 50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month Neither ships a desktop or computer-use API. On a Jarvislabs VM you install one yourself; on Baseten you add a desktop and a VNC web client to a custom sandbox image and open it through a preview URL. boat streams a 60fps desktop with Chromium and computer use. | 194th of 367, with limits $377 a month | 314th of 367, with limits $1,799 a month | 7th cheapest of 367 $200 a month |
| Browser-use agents 50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month | 261st of 356, with limits $205 a month | 331st of 356, with limits $1,194 a month | 7th cheapest of 356 $200 a month |
| Scraping & bots (startup scale) 50 at once · 1 vCPU / 2 GB · 100,000 sessions of 3 min a month | 290th of 343, with limits $247 a month | 275th of 343, with limits $183 a month | 6th cheapest of 343 $90 a month |
| Mass scraping & bots (undetectable) 1,000 at once · 1 vCPU / 2 GB · 3,000,000 sessions of 2 min a month | 156th of 327, with limits $4,942 a month | 128th of 327, with limits $3,654 a month | 73rd of 327, with limits $1,800 a month |
| Parallel red teaming 200 at once · 2 vCPU / 4 GB · 5,000 sessions of 30 min a month | 202nd of 351, with limits $127 a month | 275th of 351, with limits $381 a month | 3rd cheapest of 351 $200 a month |
| GPU eval jobs 10 at once · 8 vCPU / 32 GB · 500 sessions of 30 min a month Both fit GPU eval jobs. Jarvislabs ranks 10th at $876 a month, renting GPU VMs from L4 to H200 by the minute. Baseten ranks 21st at $1,625, on GPUs from T4 to B200. boat has no GPUs. | 10th cheapest of 102 $876 a month | 25th cheapest of 102 $1,625 a month | Not ranked: misses a requirement |
| macOS agents 5 at once · 4 vCPU / 8 GB · 1,000 sessions of 20 min a month | Not ranked: misses a requirement | Not ranked: misses a requirement | Not ranked: misses a requirement |
Key features
| Feature | Jarvislabs | Baseten | boat |
|---|---|---|---|
| Kind of product | GPU cloud: GPU VMs, GPU containers, CPU VMs, clusters, shared filesystems and model serving | Inference cloud for serving and training models; agent Sandboxes launched November 5, 2026, CLI labeled pre-release | Sandboxes for AI agents: VMs with agents, desktop and hosting ready |
| Isolation | GPU VMs and CPU VMs are KVM virtual machines; Templates are containers | Sandboxes: a microVM per sandbox with its own kernel, per launch coverage (Baseten's docs do not name it). Model deployments: containers | Every sandbox is a full VM with its own kernel |
| Root and Docker | Root on VMs, so you can run your own Docker daemon; nested virtualization is not documented | Model deployments run as the non-root app user; the sandbox docs do not document root, Docker or /dev/kvm | Docker, root, systemd and /dev/kvm on every sandbox |
| GPUs | Nvidia L4 up to H200, 1 to 8 per instance, on-demand, spot or reserved, plus clusters | Nvidia T4 up to B200, 1 to 8 per instance, multi-node over InfiniBand | None |
| CPU machines | CPU VMs from 2 to 32 vCPU with 4 GB of RAM per vCPU, in Noida only | CPU instances from 1 vCPU / 2 GiB to 16 vCPU / 64 GiB, shared, no region choice; sandboxes 1 to 32 vCPU | 4 vCPU / 8 GB by default, up to 16 vCPU / 32 GB on the $100 plan and up |
| Files after a stop | A paused VM keeps its whole disk, which keeps billing at $0.10/GB-month with at least 100 GB; shared filesystems survive pause and destroy | Model replicas start fresh from the image. Sandbox standby keeps files, memory and processes and bills only storage; sandboxes expire after 7 days by default (configurable); volumes keep files after deletion | No compute charge while stopped; the disk is saved every minute and at each stop, and kept at least 14 days |
| Snapshots and fork | No snapshot, clone or fork command; shared filesystems attach to new instances, and startup scripts or your own image rebuild a machine | No fork documented; new replicas and sandboxes start from the image | Disk saved every minute and on stop; fork a running sandbox from its last 1-minute snapshot |
| Scale to zero and wake on request | Serverless model workers stop after an idle timeout and start on the next request | Deployments scale to zero and start on the next request; standby sandboxes resume with memory and processes on the next request | No wake on request; you resume a stopped sandbox through the API, and a resume is like a reboot |
| Desktop and computer use | No desktop shipped; JupyterLab, VS Code in the browser and SSH; you can install a desktop on a VM | None documented; you can add a desktop and a VNC web client to a custom sandbox image and open it through a preview URL | 60fps desktop and Chromium, computer use and browser control, embeddable stream |
| Coding agents | A CLI and skill file so an agent on your own machine can create GPUs, run jobs and read logs | Baseten Switch routes Claude Code or Codex to its Model APIs; for sandboxes you install the agent in a custom image or drive it through the CLI or MCP endpoint | Seven coding-agent harnesses preinstalled, including Claude Code, Codex and OpenCode, driven from outside with a prompt, steer and events API |
| Reaching the machine | VMs get a public IP with every port open by default; Templates expose chosen HTTP ports through proxy URLs | HTTPS API endpoint per deployment; sandbox previews give public or private HTTPS URLs per port, plus a /port/ route with a token | HTTPS URLs for up to 50 ports, token-protected by default; SSH; raw ports over public IPv6; no dedicated IPv4 |
| Network rules | Security groups with inbound and outbound rules by port and IPv4 range (IN1 and IN2), VPCs and reserved IPs | Deployments: TCP ports 443 and 80 out by default; per-deployment IP and domain allowlists once enabled; not documented for sandboxes | Open internet on every plan; no provider-side allowlist |
| Keeping keys out of the machine | Keys are passed in as environment variables or uploaded files | Secrets are passed into the model as arguments or files under /secrets | Optional paid AI credentials proxy ($1 a day while on, plus $1 per 1,000 requests); other secrets still enter the sandbox |
| Regions | India only for new instances (IN1, IN2) | US and EU for GPU deployments; self-hosted in your own cloud account on Enterprise | Germany, Finland and France; you cannot pick one; no US or Asia region |
Every feature Battleships tracks (43)
| Feature | Jarvislabs | Baseten | boat |
|---|---|---|---|
| Full VM (own kernel) | Yes | No documented way | Yes |
| Docker inside | Yes | No documented way | Yes |
| Nested virtualization | Not published | No documented way | Yes |
| Root | Yes | Not published | Yes |
| Systemd | Not published | Not published | Yes |
| Desktop GUI | No documented way | No documented way | Yes |
| Browser | No documented way | No documented way | Yes |
| Computer-use API | No documented way | No documented way | Yes |
| SSH | Yes | Yes | Yes |
| Persistent disk | Yes | No documented way | Yes |
| Pause/resume | Yes | No documented way | Yes |
| Automatic snapshots | No documented way | No documented way | Yes |
| Fork/clone | No documented way | No documented way | Yes |
| Live fork (no pause) | No documented way | No documented way | Yes |
| Memory fork | No documented way | No documented way | No documented way |
| Custom image (Docker or snapshot) | Yes | Yes | Yes |
| ≥24 h sessions | Yes | Yes | Yes |
| Idle auto-stop | Yes | Yes | No documented way |
| Wake on request | Yes | Yes | No documented way |
| HTTPS preview URLs | Yes | Yes | Yes |
| Raw TCP inbound | Yes | Not published | Yes |
| Public IPv4 | Yes | No documented way | No documented way |
| Egress allowlist | Yes | Yes | No documented way |
| Secret proxy | No documented way | No documented way | Yes |
| Secret proxy for any API | No documented way | No documented way | Yes |
| Preinstalled agents | No documented way | No documented way | Yes |
| Agent harness API | No documented way | No documented way | Yes |
| MCP server | No documented way | Yes | No documented way |
| GPU | Yes | Yes | No documented way |
| Windows | No documented way | No documented way | No documented way |
| Arm64 | No documented way | No documented way | No documented way |
| Self-hosting | No documented way | Yes | No documented way |
| BYOC | No documented way | Yes | No documented way |
| Open source | No documented way | No documented way | No documented way |
| Scoped API keys | Not published | Not published | Yes |
| Spend limits | Yes | No documented way | Yes |
| Usage per sandbox (API) | Yes | Yes | Yes |
| Audit logs | Not published | Not published | No documented way |
| SSO | No documented way | Yes | No documented way |
| SOC 2 | No documented way | Yes | No documented way |
| HIPAA | No documented way | Yes | No documented way |
| EU data residency | No documented way | Yes | No documented way |
| Zero data retention | Not published | Not published | Yes |
Performance
Jarvislabs and Baseten are not in the public benchmarks we use, and neither publishes a CPU benchmark, so we cannot say which is faster. In StarSling's run of 3 October 2026 (run 37073680030), boat did 28.06 runs/s on the V8 Web Tooling Benchmark at 4 vCPU / 8 GB, second of 15. Every boat sandbox in this run ran on an AMD Ryzen 9 9950X host. boat also places some sandboxes on AMD Ryzen 7 9700X hosts, which this run did not measure.
V8 Web Tooling Benchmark at a 4 vCPU / 8 GB target, runs per second, higher is better. Source: StarSling's run of 3 October 2026 (run 37073680030), median of 5 sandboxes. Providers ran on different hardware in different regions. StarSling gives a shared rank when the run cannot separate two providers. Every boat sandbox in this run ran on an AMD Ryzen 9 9950X host. boat also places some sandboxes on AMD Ryzen 7 9700X hosts, which this run did not measure. Jarvislabs and Baseten are not in the public benchmarks we use, so we cannot say which is faster.
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. Start limits, per-start fees and plan minimums are in the price, boat's included.
| Scenario | Jarvislabs, a month | Baseten, 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 | $21 · On-demand | $34 · Basic · with limits | $20 · $20/mo |
| One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month | $71 · On-demand | $136 · Basic · with limits | $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 | $892 · On-demand | $1,701 · Basic · 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 | $231 · On-demand · with limits | $1,246 · Basic · 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 | $514 · On-demand · with limits | $6,228 · Basic · with limits | $8,800 · $2000/mo · with limits |
Jarvislabs: Jarvislabs bills per minute from a prepaid wallet with no plan fee and no free trial. CPU VMs cost $0.012 per vCPU-hour plus $0.0032 per GB of RAM, and disk is $0.10/GB-month while running and while paused, with at least 100 GB on a VM.
Baseten: Baseten bills per minute, rounded up, from image build and cold start until the replica stops; a 4 vCPU / 16 GiB CPU instance is $0.00346 a minute. There is no plan fee on Basic. Sandboxes are free until November 16, 2026, and pricing after that is not published, so the scenarios use CPU instances.
Baseten: Plans cost $20, $100, $500 or $2,000 a month, and the whole plan price comes back as VM time for that month. A 4 vCPU / 8 GB sandbox costs $0.036 an hour, billed per second while it runs. A stopped sandbox has no compute charge, and boat keeps its disk for at least 14 days; extra saved snapshots cost $0.056 a day each, and the optional AI credentials proxy still charges. The $20 plan includes 12 starts a minute, 60 an hour and 200 a day; extra starts cost $0.03 to $0.07 each.
Limits in those scenarios
Baseten's CPU instances have no persistent disk, so files are lost when a replica stops, and the table marks them as a compromise. Baseten Sandboxes run 10 per organization at once by default, and support raises it. Jarvislabs fits the dev box and CI scenarios, but does not publish how many machines can run at once, so 300 sandboxes for your users or 1,000 short jobs at once are not confirmed until you ask. On boat the 1,000-job burst hits the start limits: the $2,000 plan includes 90 starts a minute and 600 an hour, extra starts cost $0.03 to $0.07 each, and starts above 270 a minute fail.
| Scenario | Jarvislabs | Baseten | boat |
|---|---|---|---|
| One coding agent, workdays | None for this scenario |
| None for this scenario |
| One agent that never stops | None for this scenario |
| None for this scenario |
| 50 agents in parallel |
|
| None for this scenario |
| Sandboxes for your users |
|
| None for this scenario |
| 1,000 short jobs at once |
|
|
|
Which one should you use?
Use Jarvislabs when
you want to rent whole GPU or CPU machines with root by the minute, keep their disk between sessions, or run in India.
- A whole machine with root. Useful if you want to log in and work: Jarvislabs rents KVM VMs with root, SSH and a disk that survives a pause.
- Big CPU VMs by the minute. Useful for heavy CPU work: Jarvislabs CPU VMs go up to 32 vCPU / 128 GB, billed per minute with no plan fee.
- Firewall rules and India regions. Useful for teams in India, or who want security groups, VPCs and reserved IPs set by the provider.
Use Baseten when
you serve or train models in production and want deployments that scale to zero, sandboxes that keep memory on standby, GPUs up to B200, or the US or your own cloud account.
- Scale to zero and standby with memory. Useful for spiky traffic and idle agents: deployments scale to zero, and standby sandboxes resume with memory and processes on the next request.
- GPUs up to B200, across nodes. Useful for large models: Baseten runs Nvidia GPUs from T4 to B200, up to 8 per instance and across nodes.
- US region or your own cloud. Useful if you must run in the US, or in your own AWS, GCP or OCI account on Enterprise.
Where boat fits
Jarvislabs and Baseten are GPU clouds, and boat has no GPUs, so keep model training and serving on one of them. If you run agents that write code, use a browser or host previews, boat gives each one a full VM built for that. Many teams split it: models on a GPU cloud, agents on boat. boat's limits: start caps per plan, and no US or Asia region.
A disk saved every minute, and fork
A paused Jarvislabs VM keeps billing for at least 100 GB of disk, and copying a set-up machine means a new instance plus a startup script or shared filesystem. Baseten sandboxes keep files on standby, but its docs show no fork. boat saves the disk every minute and forks a running sandbox; if a machine fails, you lose what was written after the last good snapshot.
Agents and a desktop already inside
Jarvislabs and Baseten give an agent on your laptop tools to manage machines or call models, not a machine with the agent in it. On both you install the agent, a browser and any desktop stream yourself.
Token-gated HTTPS ports and model keys kept off the machine
A Jarvislabs VM opens every port until you add a security group. On both, real API keys sit on the machine, where an agent can leak them in a log. boat URLs carry a token by default, and its optional paid credentials proxy keeps model-provider keys outside.
Use boat when you run agents: full VMs with root and Docker, coding agents and a 60fps desktop inside, token-gated HTTPS on 50 ports, a disk saved every minute, live fork, and no compute charge while stopped. Compare it one to one: boat vs Jarvislabs and boat vs Baseten.
boat's own limits
- Plans start at $20 a month. The plan price comes back as VM time for that month. Unused time does not carry over. Source
- Starts (create, fork, resume): the $20 plan includes 12 a minute, 60 an hour and 200 a day. The $2,000 plan includes 90 a minute, 600 an hour and 2,400 a day. Extra starts cost $0.03 to $0.07. A start fails with 429 above 3 times the included starts per minute, or 6 times per hour or per day. Source
- A stopped sandbox has no compute charge. boat keeps its disk for at least 14 days. A saved snapshot keeps it longer; extra saved snapshots and the AI credentials proxy are paid. Source
- Auto-stop defaults to 1 hour. You can turn it off on a paid plan. The trial allows 2 hours at most. Source
- boat saves the disk every minute and at each stop. Memory, running programs and /tmp are not saved: a resume is like a reboot. A machine failure loses writes after the last good snapshot. Source
- Machines run in Germany, Finland and France only. Snapshot copies can be stored outside the EU. Source
- Public IPv6 and outbound IPv4. No dedicated IPv4. The IP changes at each resume. Source
- Hosts are AMD Ryzen 7 9700X and Ryzen 9 9950X. vCPUs are shared. No GPU. You cannot boot your own Docker image. Source
Questions people ask
Is Jarvislabs cheaper than Baseten?
In the CPU scenarios we priced, yes. A 24/7 dev box is $68 a month on Jarvislabs and $130 on Baseten CPU instances, and CI runners are $253 against $1,119. Baseten is a compromise in both, and its Sandboxes have no published price after November 16, 2026. On boat those are $26 and $86. For GPU eval jobs Jarvislabs is also lower, $876 against $1,625.
Which is faster, Jarvislabs or Baseten?
Neither is in the public benchmarks we use, and neither publishes a CPU benchmark, so we cannot say. In StarSling's run of 3 October 2026, boat did 28.06 V8 Web Tooling runs/s at 4 vCPU / 8 GB, on a Ryzen 9 9950X host.
Can Jarvislabs or Baseten run Docker?
On Jarvislabs VMs you have root, so you can run your own Docker daemon. Baseten's model containers run as the non-root app user, and its sandbox docs do not document root or Docker inside; it can run your own Docker image as a model server. Every boat sandbox has root and Docker.
Do Jarvislabs and Baseten keep files when a machine stops?
A paused Jarvislabs VM keeps its whole disk and keeps paying for it, its GPU containers keep only /home, and shared filesystems survive pause and destroy. Baseten replicas start fresh from the image. A Baseten sandbox on standby keeps its files, expires after 7 days by default (configurable), and volumes keep files after deletion. A stopped boat sandbox has no compute charge and keeps its disk for at least 14 days.
Where do Jarvislabs and Baseten run?
Jarvislabs creates new instances only in India, and its old EU1 region no longer accepts new ones. Baseten runs GPU deployments in the US or the EU, and Enterprise customers can self-host in their own cloud. boat sandboxes run on machines in Germany, Finland and France, and you cannot pick one.
Does boat have GPUs?
No. boat runs Linux VMs on AMD Ryzen hosts and has no GPUs. Train or serve models on Jarvislabs or Baseten, and let your agents work on boat.
Try boat
Install the CLI, then run boat new.
7-day free trial