Pricing

Pay for the platform, not for tokens

almyty is open source, so self-hosting is free and always will be. Hosted plans are metered on agent runs, not tokens — and you choose how inference is billed: your own keys with no markup, or managed by almyty. No lock-in either way.

Free

$0

Hosted, to get going

Enough to build and run real agents on your own keys.

  • Bring your own LLM keys
  • 1 user, generous run allowance
  • MCP, A2A, UTCP, and Agent Skills gateways
  • Community support
  • Self-host the open source for free, forever
Most popular

Pro

$20/ user / month

For builders shipping agents

Unlimited resources, metered on the runs you actually make.

  • Everything in Free
  • Large monthly run allowance, then pay per run
  • Unlimited gateways, tools, and agents
  • Scheduling, webhooks, analytics, audit log
  • Email support

Business

$60/ user / month

For teams in production

The controls and scale a team needs to run agents for real.

  • Everything in Pro
  • Higher run allowances and rate limits
  • SSO and role-based access control
  • Managed inference option
  • Priority support

Enterprise

Custom

Call us

For security, scale, and deployment requirements.

  • Volume commitments and custom limits
  • Self-host, private cloud, or BYO-Bedrock/Vertex
  • SSO/SCIM, SLAs, and security review
  • Dedicated support and onboarding
  • Open source — take it with you

Run allowances and per-run rates are shown in the app. Annual billing lowers the per-seat price.

Inference, your way

Bring your own key, or let us manage it

Pick whichever you prefer — and mix them per agent. Switch at any time; nothing is locked to one choice.

No markup

Bring your own key

Connect your own provider keys — OpenAI, Anthropic, Bedrock, Vertex, and the rest. almyty adds nothing to the token cost; you pay the provider directly and keep your own rates and quotas.

Pay as you go

Managed tokens

Skip key management and let almyty handle inference. You pay for tokens at provider cost plus a small platform margin, billed with the rest of your usage. Nothing to set up.

Mix BYOK and managed inference across agents in the same workspace.

Not sure which fits?

Start on the free tier or self-host today. Move up when you need scale, SSO, or a private deployment. Nothing you build is trapped on one plan.