Mix managed runners with your own in one workspace. Pay each runner’s rate for the time it works. No seat fees, no monthly minimum, and no markup on model tokens. Your data remains yours.
Entergenie operates the platform, standard runners, and connections for you.
Your runners and connections in your VPC, with the same cloud platform as managed runners.
Fully self-hosted for organizations that need the entire platform inside their own infrastructure.
All three include the same agent, task, policy, approval, and audit capabilities. Fully managed and your runners share one cloud platform: combine them across tasks and agents without a plan change. Enterprise runs the whole platform in your environment. Model usage is separate. Larger or GPU runners, dedicated capacity, extended retention, and custom engineering require an agreed quote before use.
Use managed runners where convenience matters and your own runners where the work needs to stay in your VPC. Add both kinds of usage together.
10,000 combined active runner minutes per month. Excludes model usage, taxes, and infrastructure you host. Standard managed compute is included. No minimum or fee to mix runner types.
Hosting the entire platform yourself? Enterprise is a separate, fully self-hosted deployment at $0.002/min. The same 10,000 total minutes would be $20.00 in software usage, plus your infrastructure and models.
| Responsibility | Fully managed | Your runners | Enterprise · Fully self-hosted |
|---|---|---|---|
| Platform & task records | Entergenie hosts | Entergenie hosts | You host |
| Agent runners / VMs | Entergenie hosts | You host | You host |
| Connections & credentials | Entergenie hosts | You host | You host |
| Infrastructure charges | Standard runner compute included | Runner infrastructure paid to your provider | Paid directly to your infrastructure provider |
| LLM usage | Provider cost, no markup; or bring your own credentials | Pay your provider directly | Pay your provider directly |
| Data ownership | Yours | Yours | Yours |
An active runner-minute is 60 seconds of a runner allocated to executing task work. We total seconds across your runners each month, then round the charge once to the nearest cent.
Billing starts when task execution begins and stops when execution pauses, stops, or completes. Queue time, runner provisioning, and paused approval waits are free. Waiting for a model or tool response during execution counts.
Two runners executing for 30 seconds use one runner-minute in total. Splitting a task into sessions adds no fee. Viewing tasks or opening the UI is free.
Entergenie execution charges for platform-caused failures are credited. Task errors, external connection failures, and user-requested retries count for the execution time they consume.
Choose task dollar budgets, time limits, and a monthly Entergenie usage cap. Pause new execution when the remaining budget cannot cover it. Model-provider budgets are managed separately.
No. Model usage is separate from runner time. With fully self-hosted or customer-hosted runners, bring your own credentials and pay your provider directly. With fully managed, use your own credentials or have Entergenie pass through the model provider’s actual usage cost with no markup, itemized separately. We never add a percentage to token costs.
No. There are no seat, agent, task, session, or connection fees. Normal platform features are included in the runner-minute rate. External services can still charge their own fees. Dedicated infrastructure, unusual resource requirements, custom connectors, and hands-on implementation are agreed separately before work begins.
Entergenie operates the platform, standard 2 vCPU / 4 GiB runners, and standard connections. The $0.01 per active runner-minute includes standard compute, runtime storage, and networking. Model usage is separate. Persistent bulk storage, extended retention, larger or GPU runners, and dedicated environments are quoted separately and require approval; they are not automatic overages.
Fully self-hosted means your team operates the platform, task records, runners, and connections in your infrastructure. Billing requires only aggregate execution seconds and a deployment identifier, not prompts, outputs, credentials, or source data. A disconnected deployment can submit a monthly usage summary. If you choose an external LLM API, the context you send still leaves your environment.
Yes. Use fully managed and customer-hosted runners in the same cloud workspace. Select the runner environment for your agents and sessions. Each execution is billed at its own rate: $0.01/min managed or $0.004/min on your runners. For example, 5,000 managed minutes plus 5,000 minutes on your runners cost $70 in Entergenie usage. There is no plan-switching fee or duplicate platform subscription. A runner must have approved access to the connections needed for its task.
Choose Enterprise when the entire platform, task records, runners, and connections need to run in infrastructure your organization operates. It is the fully self-hosted offering, with deployment architecture, identity, network boundaries, and operating ownership agreed for your organization. The software usage rate remains $0.002 per active runner-minute; infrastructure and models are paid directly by you. Optional implementation and dedicated support are quoted separately.
With your runners, Entergenie hosts the coordination platform while your VMs and connections run in your VPC. Agree which task context, results, and audit records reach the cloud platform. With fully self-hosted Enterprise, you operate that platform too. You can mix your runners with managed runners on the cloud platform; moving the platform itself into your infrastructure is an Enterprise deployment. In both cases, you pay your own infrastructure and model providers directly.
Set a monthly Entergenie execution-spend cap, alongside task budgets and time limits. New work pauses when the remaining allowance cannot cover it; raising a cap is an explicit decision. Model spend and infrastructure billed by your own providers need their own limits. Entergenie-arranged model usage has a separate agreed cap and is itemized at cost.
There is no mandatory setup fee, prepaid usage bundle, monthly minimum, or annual commitment. Standard usage is invoiced monthly in arrears. Optional migration, custom integrations, training, dedicated support, and implementation services receive a separate quote that you approve first. You can stop future usage at any time and settle the usage already incurred.
Yes. Each execution period is charged at the deployment rate in effect for that runner, so moving between deployment options does not reprice previous usage. Plan data export, connection access, and the operating handover before moving. Your business data remains yours.