# Server4Agent

Official pricing: https://www.server4agent.com/pricing  
Category: dev-env · Isolation: container

## Pricing regimes (raw)

- **Native product meter; not a general-purpose 4-vCPU/8-GiB VM** (sizes):  [alt]

## Features

Yes: snapshots, pause/resume, persistent disk, volumes, idle auto-stop, ≥24 h sessions, custom image (Docker or snapshot), start from your own snapshot, HTTPS preview URLs, open internet, BYOC, Python, Node.js, wake on request, webhooks, EU data residency, scoped API keys, spend limits, MCP server, agent harness API, hosted agent API (their own agent), inbound access rules

No: memory snapshots, fork/clone, your own Docker/OCI image, full VM (own kernel), Docker inside, nested virtualization, browser, desktop GUI, computer-use API, browser + desktop control, anti-bot stealth, CAPTCHA solving, residential IPs, preinstalled agents, SSH, public IPv4, egress allowlist, static egress IP, self-hosting, open source, SOC 2, HIPAA, SSO, GPU, arm64, Windows, macOS, HTTP method/path egress rules, live resize, live fork (no pause), memory fork, audit logs, automatic snapshots, snapshots on demand, gVisor or VM (no shared kernel), firewall inside (nftables), secret proxy, secret proxy for any API, extra volumes, shared volumes

Unknown: everything else. Evidence (source + quote) per feature: https://battleships.dev/data/providers/server4agent.json → feature_evidence

## Caveats

- vCPU/RAM of server tiers and base compute meter unpublished.
- Plan fee is not wholly credit: Starter $29 includes $10; Pro $99 includes $40; Business $299 includes $150.
- Displayed ~1.4x Starter/Pro and ~1.2x Business AI overage are approximate and applied to an unpublished base; cannot derive a unit price.
- Default managed-server caps stop at 10, so 50 concurrent servers requires Enterprise contract.
- Project count is not sandbox concurrency; no inference that many projects each get a separate 4/8 VM.
- Early-access CTA leaves onboarding availability untested; no account created.
- Native billing cannot be reduced to a 4-vCPU/8-GiB price. No matching engine sizes, by design.
- Included credits reset every billing period (no rollover). Free plan hard-stops; paid plans can allow overage (~1.4x base on Starter/Pro, ~1.2x on Business) or keep a strict cap. Budget caps can be slightly exceeded because of propagation delays.
- Product is still early access: every plan CTA is a waitlist button and /waitlist redirects to /?waitlist=1; self-serve signup not confirmed.
- Annual billing gives 2 months free: Starter $290/yr (~$24/mo), Pro $990/yr (~$83/mo), Business $2,990/yr (~$249/mo). Not encoded as plans.
- Included credits may cover only AI/agent work: a Server4Agent blog says server tiers (small/medium/large) and egress are metered separately and 'priced by capacity' (no tier price published), while the pricing FAQ says a credit is '$1 of metered usage (compute + AI)'. Treat included_usd as applying to AI usage only until clarified.

## How this provider charges


As of 2026-09-28. Public USD list subscription allowances, plus metered dollars of compute+AI. Signup/early-access product with published MCP and REST; original launch date not verified. The provider explicitly distinguishes itself from raw sandbox primitives.

## Regimes

| Regime | When it applies | How billed / numbers | Source |
|---|---|---|---|
| Account credits | Native product usage | 1 credit=$1 of compute+AI metered usage; base meter not published. | [Primary pricing](https://www.server4agent.com/pricing) |
| Egress overage | Native product usage | Starter $0.12/GB after 100 GB; Pro $0.10/GB after 250 GB; Business $0.08/GB after 1 TB; Free 10 GB with no published overage. | [Primary pricing](https://www.server4agent.com/pricing) |
| Annual discount | Native product usage | Two months free advertised; credit/renewal details not confirmed. | [Primary pricing](https://www.server4agent.com/pricing) |
| Free | Account plan | $0/month; 1 server; small tier; 5 GB/server; 1 public project; 10 GB monthly egress; $2 trial credits. | [Primary pricing](https://www.server4agent.com/pricing) |
| Starter | Account plan | $29/month; 2 small servers; 5 GB/server; 3 public projects; 100 GB egress; 3 members. | [Primary pricing](https://www.server4agent.com/pricing) |
| Pro | Account plan | $99/month; 3 small/medium servers; 20 GB/server; 10 public projects; 250 GB egress; 5 members. | [Primary pricing](https://www.server4agent.com/pricing) |
| Business | Account plan | $299/month; 10 servers, small/medium/large; 100 GB/server; 30 public projects; 1 TB egress; 15 members. | [Primary pricing](https://www.server4agent.com/pricing) |
| Enterprise | Account plan | Unpublished / sales; Unlimited advertised servers, custom storage/credits; managed or BYOC. | [Primary pricing](https://www.server4agent.com/pricing) |

## Gotchas

- vCPU/RAM of server tiers and base compute meter unpublished.
- Plan fee is not wholly credit: Starter $29 includes $10; Pro $99 includes $40; Business $299 includes $150.
- Displayed ~1.4x Starter/Pro and ~1.2x Business AI overage are approximate and applied to an unpublished base; cannot derive a unit price.
- Default managed-server caps stop at 10, so 50 concurrent servers requires Enterprise contract.
- Project count is not sandbox concurrency; no inference that many projects each get a separate 4/8 VM.
- Early-access CTA leaves onboarding availability untested; no account created.

## Required worked example

4 vCPU / 8 GiB, 50 concurrent × 8 h/day × 22 days = **8,800 instance-hours**, 30% CPU, 50 GiB retained snapshots and 100 GiB egress.

**Enterprise quote required.** 50 simultaneous servers exceeds Business cap 10; vCPU/RAM and compute pricing unknown. 100 GiB egress would fit Starter/Pro/Business allowances under the project convention GB≈GiB, but that does not solve the capacity mismatch or price the retained snapshots.

## Feature evidence

- https://www.server4agent.com/llms-full.txt
- https://www.server4agent.com/docs/api
- https://www.server4agent.com/blog/e2b-alternatives-for-ai-agents