You pay for work done. Building your system is included.

Panaptic's agent builds your integrations, code libraries, and apps as part of the platform — there's no separate build bill. Two axes meter what runs: Platform Units for storage and engagement, AI Credits as a pay-as-you-go consumption charge. Manual mode uses zero AI Credits, so a human driving the work costs nothing in model spend.

Contact us

Starter

For solo operators describing their first apps and workflows.

$29/mo

  • ✓ $29 of AI Credits included
  • ✓ 1 GB workspace storage
  • ✓ Up to 100 task runs/day
  • ✓ Manual & CoPilot execution modes
  • $ Save ~17% with annual billing — $289/yr
Most popular

Professional

For small teams running real workloads and automation.

$200/mo

  • ✓ $200 of AI Credits included
  • ✓ 20 GB workspace storage
  • ✓ Up to 5,000 task runs/day
  • ✚ Manual, Co-pilot & Autonomous execution modes
  • ✚ Webhook & scheduled triggers
  • $ Save ~17% with annual billing — $1,989/yr

Business

For growing teams with large workloads and public apps.

$600/mo

  • ✓ $600 of AI Credits included
  • ✓ 75 GB workspace storage
  • ✓ Up to 10,000 task runs/day
  • ✓ Manual, Co-pilot & Autonomous execution modes
  • ✓ Webhook & scheduled triggers
  • ✚ Public task deployment
  • $ Save ~17% with annual billing — $5,979/yr

Scale

For large teams with high workloads and throughput.

$1,200/mo

  • ✓ $1,200 of AI Credits included
  • ✓ 200 GB workspace storage
  • ✓ Up to 100,000 task runs/day
  • ✓ Manual, Co-pilot & Autonomous execution modes
  • ✓ Webhook & scheduled triggers
  • ✓ Public task deployment
  • ✚ Priority support
  • $ Save ~17% with annual billing — $11,949/yr

Need more than Scale — single-tenant isolation, custom SLAs, or a compliance pack? Talk to us →

How billing works

Two axes. No build bill. Cost you can actually predict.

Platform Units

Your margin layer, metered by storage and engagement — tasks, data, runs, retention. The agent authoring your integrations, code libraries, and apps is part of the platform, so building costs nothing extra; you pay as the system you described starts carrying real work. Per-org PostgreSQL, tenant isolation, and OpenTelemetry export are included, not held back for an enterprise tier.

AI Credits

A pay-as-you-go consumption charge - you're billed for the model usage your org actually consumes. Manual mode uses zero AI Credits: when a human drives the task and the agent only assists on request, you spend nothing on model calls. CoPilot and Autonomous draw credits as the agent proposes or executes, weighted by the model you've chosen for your org.

No upsell tax

The hard infrastructure ships on every plan, not behind a sales call.

Per-org Postgres

Each org gets its own PostgreSQL database, with tenant boundaries enforced at the storage layer — on Starter and up.

Audit log

Every action is signed by its actor — user, agent, or system — with operation, inputs, and timestamp. The audit log is the kernel, always on and queryable.

Replay-safe execution

Execution re-runs from the top on every integration boundary with deterministic params, so non-idempotent operations can't fire twice on a retry.

Schema-level access rules

Scope what each role and agent can call, per task, defined against the same JSON Schema contracts that type the rest of the platform.

Model-agnostic agents

Choose your LLM provider per org — Claude, Gemini, others. Tools, not chains: no LangChain or LangGraph lock-in.

Live authoring

The agent authors integrations, typed code libraries, apps, and data tables inside the platform — verified against real responses, no separate codebase, no deploy step.

Not sure which plan fits?

Begin with Starter, describe how your business works, and watch the agent build your first task in minutes — or talk to us about a guided rollout for your team.

Contact us