Skip to main content

​
Plans

Four self-serve tiers. Every plan includes container-isolated agent execution, 100+ LLM providers via LiteLLM, the OpenLegion credit-backed default model, monitoring dashboard, custom subdomain, multi-channel support (Telegram / Discord / Slack / WhatsApp / Webhook), and a dedicated single-tenant Hetzner Cloud VPS (ARM preferred, x86 fallback).
BasicGrowthProPro Max
Monthly$19/mo$59/mo$149/mo$279/mo
Yearly$170/yr$530/yr$1,340/yr$2,510/yr
Agents151530
Browsers151030
Projects02510
SupportCommunityCommunityCommunityCommunity
Yearly billing saves approximately 25%.
Pro is the most popular plan — 15 agents and 5 projects covers most production use cases.
No free trial. Your card is charged at checkout. Payments are processed by Polar.

​
Enterprise

For higher limits, dedicated infrastructure, custom contracts, or contractual support, contact admin@openlegion.ai. Enterprise is not a self-serve tier.

​
LLM usage and credits

Every managed instance is provisioned with OpenLegion’s credit-backed proxy as its default model (openlegion/openai/gpt-5.4). When your subscription first becomes active, welcome credits are granted automatically. You can also:
  • Bring your own API keys (BYOK) for any of the 11 supported providers — they overlay on top of the credit proxy, giving you access to additional models.
  • Top up credits from the /credits page in the app.
Engine-side, every LLM call is charged to a per-agent + per-project ledger (data/costs.db). Defaults: $10/day and $200/month per agent. The operator agent can raise these up to $1,000/day and $30,000/month. CAPTCHA-solving spend is tracked in a separate ledger in millicents (1/100,000 USD).

​
Changing plans

Manage your plan from the Settings page at app.openlegion.ai.

​
Upgrades

Upgrades take effect immediately. You’re charged a prorated amount for the remainder of your billing period. Your agent, project, and browser limits increase right away. If the upgrade requires a larger server, an automatic resize happens. Resize sequence: pre-scale limits → power off → change_server_type → power on → wait for SSH → re-apply limits → wait for engine health (up to 240s). Total downtime: ~30-60 seconds.

​
Downgrades

Downgrades are scheduled for the end of your current billing period. You keep your current plan’s limits until then. At renewal, your plan switches and limits adjust. You can cancel a pending downgrade from the Settings page by clicking Keep current plan.

​
Billing interval changes

Switching between monthly and yearly billing takes effect at your next renewal.

​
Cancellation

Cancel from the Settings page. Your instance stays active until the end of your current billing period (currentPeriodEnd). After that, the server is deprovisioned and data is deleted. You can resume a cancelled subscription at any time before the billing period ends. There is no data export UI — back up anything you need from the engine dashboard before the period ends.

​
Payment issues

If a payment fails, your subscription enters a past_due state. The app considers your subscription active during the grace period.
  • Day 0-3 — banner shown in the app; instance keeps running. Update payment to clear.
  • Day 3 — instance is paused (systemctl stop via provisioner). The VPS is preserved; resume is automatic once payment succeeds.
  • Day 14 — instance is permanently deprovisioned. The VPS and all data are deleted.
Authentication for the past-due cron uses a timing-safe bearer check; the provisioner is pinged first and the action is skipped if it’s unreachable.