Intelligence Credits — Complete Reference
Your monthly plan covers everyday governance work. Intelligence Credits cover the high-value AI outcomes that would otherwise take consultants weeks to produce. 1 Credit = $1.00 at the base rate. You always see the cost before a job starts.
The Credit Lifecycle
Before every credit-bearing job starts, Arcitopsia displays an estimated credit range based on the scope you have configured — the number of components, complexity of your architecture, and your plan tier. You decide whether to proceed.
The estimate shows a minimum and maximum credit range. The actual charge falls within that range based on what the job actually executes.
When you confirm, the maximum estimated credits are temporarily reserved in your balance — not deducted. Your job runs with full access to your knowledge graph, standards, and registry context.
Reserved credits remain in your balance during execution. Failed jobs release the full reservation — you are never charged for a workflow that does not complete successfully.
Once the job completes, only the actual credits consumed are deducted. Any reserved-but-unused credits are immediately returned to your balance. Your credit statement shows the exact breakdown for every job.
The actual cost is always ≤ the maximum estimate. Scope that runs faster or touches fewer components than anticipated costs less.
Credit-Bearing vs. Included
Daily governance work is covered by your plan's workflow job quota. Intelligence Credits apply only to high-value outcomes that replace days or weeks of consulting work.
The rule is simple: any workflow that calls an LLM to generate new content costs credits — code, IaC, diagrams, documents, enrichment. Operations that involve no LLM generation (routing, versioning, displaying, evaluating rules) are covered by your plan.
Activity Pricing
All prices shown are at the base credit rate of $1.00 per credit. Actual credit cost varies with the scope of your specific job. You always see a range estimate before confirming. Credits are never deducted for failed jobs.
| Outcome | What Arcitopsia Produces | Credits Required | At Base Rate |
|---|---|---|---|
| Architecture Intelligence | |||
| Architecture Document | Full design document for a solution, service, or domain — including context narrative, C4 diagrams, dependency graph, ADR, and alignment to active standards. Generated from your live registry, not internet examples.Scope: 1 solution / service set. Larger programmes span multiple documents. | 800base; scales with service count & standards depth | from $800 |
| Architecture Simulation Run | Model what-if scenarios — traffic projections, infra cost changes, failover impacts, tech deprecation. AI runs against your actual dependency graph to produce a concrete analysis and recommendation.Single scenario; complex multi-variable simulations use more credits. | 200base; scales with model complexity | from $200 |
| Code & Diagram Generation | |||
| Code Generation | AI-generated code scaffolding for a component, module, or microservice — implementation stubs, interfaces, tests, and inline documentation aligned to your registry's SpecType schemas and active standards.Per component; multi-module scaffolds and complex service meshes use more credits. | 80base; scales with complexity & component count | from $80 |
| IaC Generation | AI-generated Infrastructure as Code — Terraform modules, Pulumi stacks, CDK constructs, Helm charts, or Ansible playbooks derived from your registry records and delivery team type configurations.Per module or stack; complex multi-cloud IaC with many resource types uses more credits. | 120base; scales with resource count & provider complexity | from $120 |
| Diagram Generation | AI-generated C4 diagrams, architecture diagrams, sequence diagrams, or network topology maps — built from your live registry data so they always reflect current state, not hand-drawn approximations.Per diagram set (context + container level); component and deployment diagrams count separately. | 60base; scales with diagram type & scope | from $60 |
| Runbook Generation | AI-generated operational runbooks, incident response procedures, or onboarding guides — grounded in your service registry, deployment topology, and active workflow definitions.Per runbook; incident playbooks and multi-service runbooks use more credits. | 75base; scales with runbook length & service breadth | from $75 |
| Compliance & Governance | |||
| Compliance Report | Full evidence package for SOC 2, HIPAA, PCI DSS, ISO 27001, or GDPR. Arcitopsia maps controls to your live registry, auto-collects evidence, identifies gaps, and produces a board-ready compliance report with artefact links.Single framework; multi-framework assessments require multiple reports. | 1,400base; scales with org scope & control count | from $1,400 |
| Full Compliance Assessment | Broad gap analysis across a compliance programme or domain — identifies which controls are met, partially met, or missing, with a remediation roadmap prioritised by risk. Stops short of producing a complete evidence package.Covers the assessment phase; full reports are a separate activity. | 400base; scales with domain breadth | from $400 |
| AI Agent & Autonomous Workflows | |||
| Autonomous Agent Run | A multi-step agentic workflow that executes across connected tools — Jira, Confluence, GitHub, Terraform, AWS — with real outputs: created tickets, generated IaC, updated records, sent notifications. Registry context injected automatically.Simple linear agents start lower; complex parallel agents with many tools use more credits. | from 100scales to 600 for complex parallel agents with many tools & branches | from $100 |
| Package & Data Operations | |||
| AI Enrichment Run | AI-generated descriptions, classification tags, and metadata fields for a single registry record. The LLM receives the current record's context plus a few levels of dependency records to ground the output, then generates enriched field values written back to the record — equivalent in cost to roughly one-third of a diagram generation.Priced per record. A run covers any number of records; credits reserved upfront for the full batch, settled per record on completion. | 20per record · current record + dependency context injected per call | $20 / record |
| Package Installation | Automated provisioning of a marketplace package into your tenant — creating SpecTypes, Records, LinkTypes, Workflows, AI Agents, Roles, and Connectors as defined in the package manifest. Fully idempotent; re-runs are safe.Per installation; upgrading a package to a new version counts as a new installation. | 480per package; large packages may use more | from $480 |
| Bulk Import | Large-scale registry population from CSV, JSON, or integration connector data. AI analyses the source schema once per job to establish SpecType mappings, then runs normalisation, deduplication, and link inference across the full record set. Credits are charged at $100 per 100 successfully imported records — duplicates detected and skipped are not charged.Priced per 100-record block. Credits reserved upfront for the declared batch size and settled on completion. Minimum 100 records per job. | 100per 100 records · batch settled on completion | $100 per 100 records |
| Probe Framework | |||
| Probe Discovery Run | A single scan of an external system (cloud account, source-repository organisation, database, IdP, observability backend) that ingests its inventory into your Knowledge Graph as Records and Links. Library-first extraction — most stages emit zero LLM calls. LLM only fires for ambiguous classification, ownership inference, or semantic enrichment, gated by per-stage confidence thresholds.Per ProbeRun, regardless of target count. Multi-target probes (e.g. five AWS accounts under one Organisation) fan out without multiplying the base charge — the bulk-import line covers the per-Record settlement. | 100base per ProbeRun | from $100 |
| EA Artefact AI Draft | AI-drafted Reference Architecture, Architecture Standard, Architecture Guardrail, or Architecture Pattern document, generated from observed best-of-fleet records and existing organisational patterns. Always lands in the EA Review Queue — never auto-published. Used by Hierarchy Synthesis to fill out the EA Stream when probe-discovered Delivery-Stream Records have no governing artefact.Per drafted artefact. A probe that discovers a service using Postgres, Kafka, and Stripe-SDK may draft up to four artefacts per technology (16 total), each billed independently. | 50per drafted artefact | from $50 each |
Included Credits per Plan
Included credits reset at the start of each billing period. Unused included credits do not roll over. They are designed to cover a typical month of high-value AI activity at each plan scale.
| Plan | Monthly Credits Included | Equivalent Value | Example Usage | Plan Price |
|---|---|---|---|---|
| Startup | 5,000 | $5,000 | ~3 arch docs + 10 diagrams + 15 code gens + 2 IaC modules + 1 assessment/month | $4,999/mo |
| Professional ✦ Most Popular | 15,000 | $15,000 | ~8 arch docs + 30 diagrams + 40 code gens + 10 IaC modules + 2 assessments + 1 compliance report/month | $13,999/mo ~7% off |
| Enterprise | 45,000 | $45,000 | ~20 arch docs + 100 diagrams + 100 code gens + 30 IaC modules + 2 compliance reports + 5 assessments + 30 simulations/month | $39,999/mo ~11% off |
| Strategic | 150,000 | $150,000 | ~60 arch docs + 300 diagrams + 300 code gens + 200 IaC modules + 6 compliance reports + 100 runbooks + 1,000 records enriched/month | $99,999/mo ~33% off |
| Global | 500,000 | $500,000 | ~600 arch docs or 350 compliance reports or 25,000 records enriched or 500,000 records imported/month; purchase additional credits for usage beyond included budget | Custom · annual flat fee |
| Hyperscale · 500+ users | Negotiated | Negotiated | Custom monthly credit pool sized to your actual workload; volume rates on additional purchases; no standard ceiling | Contact Support |
Plan included credits are sized to cover typical monthly AI operations at each org scale — no consumption surprises. Unused included credits expire at period end. Additional top-up credits are priced at 1 credit = $1.00 and do not expire while your account is active. See full plan details →
Additional Credits
Purchased credit packs do not expire while your account is active — they expire only after 12 consecutive months of zero credit usage. Buy larger packs to unlock volume discounts of up to 35%.
Covers 1 architecture document or 3–4 simulation runs
Covers 1 compliance report with credits to spare
Covers ~1 full compliance report + 4 architecture documents
Best value for high-volume programme delivery teams
Annual commitment — large-scale EA programmes & multi-region rollouts
| Pack | Credits | Pack Price | Rate per Credit | Discount | Expiry |
|---|---|---|---|---|---|
| Starter | 100 | $100 | $1.00 | — | 12 months inactivity |
| Growth | 500 | $450 | $0.90 | 10% off | 12 months inactivity |
| Scale | 2,000 | $1,600 | $0.80 | 20% off | 12 months inactivity |
| Enterprise | 10,000 | $6,500 | $0.65 | 35% off | 12 months inactivity |
| Strategic Annual | 50,000 | $30,000 | $0.60 | 40% off | Contract term + 10% carryover |
Inactivity expiry: purchased credits expire only if no credits are used for 12 consecutive calendar months. Any credit deduction — from included monthly credits or from a purchased pack — resets the 12-month inactivity clock.
Bring Your Own LLM
Reduced Intelligence Credit rates — the LLM inference cost component is removed from your credit price. You pay Anthropic, OpenAI, or Google directly; Arcitopsia charges only for the platform orchestration layer.
15% discount on your license fee — applied to your monthly plan subscription as recognition that you bear the LLM vendor relationship and cost.
Full data residency control — your LLM API key means your data goes to your chosen model endpoint. Useful for regulated industries requiring data residency guarantees.
Model choice flexibility — swap between Claude, GPT-4, or Gemini at any time. Use different models for different workflow types depending on performance and cost profile.
Green = fully supported today. Others: contact sales for availability.
Org Type & Flat Pricing
Credit Terms & Expiry
Every credit type has its own expiry rule. Here is the complete policy for each.
Credits bundled with your plan (e.g., 5,000 credits/mo on Startup, 45,000/mo on Enterprise) reset at the start of each billing period. Unused included credits do not roll over to the next month. This encourages active use and reflects the reserved platform capacity your subscription provides.
Credits from a purchased pack do not expire while your account is active and credits are being used. They expire only after 12 consecutive calendar months of zero credit usage. Any credit deduction from any source — included credits or purchased packs — resets the 12-month inactivity window.
Credits in an annual committed package are valid for the contract term (typically 12 months). Up to 10% of unused committed credits may be carried over to the next contract year by agreement at renewal. Carry-over must be agreed in writing before the contract term ends.
Credits provided during a free trial period expire 30 days from account activation, regardless of usage. Trial credits cannot be converted to paid credits and are not refundable. Trial credits are separate from your first month's included credits, which activate on your first paid billing cycle.
Credits are never deducted for failed jobs. When a workflow errors, times out, or is cancelled before completion, the full credit reservation is released immediately. Your balance returns to exactly what it was before the job started. This applies to all credit-bearing activities without exception.
When a job settles, credits are deducted in this order: (1) included monthly credits first, then (2) oldest purchased pack credits, then (3) more recent purchased pack credits. This maximises the life of your purchased credits and ensures included credits are consumed within the period.
FAQ
Yes. Every credit-bearing activity shows an estimated credit range (minimum–maximum) before you confirm. The range reflects the configured scope. You always see the estimate, confirm, and only then does a reservation happen. You can cancel before confirming at no cost.
If your credit balance falls below the minimum reserve needed for a job, the job will not start. Arcitopsia shows a low-balance warning before this point. You can purchase a top-up pack at any time — credits are added to your balance immediately after payment. You can also enable auto top-up on a chosen pack size when your balance falls below a threshold you set.
Yes. Intelligence Credits are held at the organisation (tenant) level, not the individual user level. Any team member with permission to run credit-bearing workflows can use the shared credit balance. Org admins can set per-user or per-team credit spending limits in the billing dashboard.
Purchased credit packs are non-refundable. If you cancel your plan, you retain access to your purchased credits for 30 days to run any final jobs. After 30 days the credits expire. Contact support if you have an exceptional circumstance — refunds are handled case by case.
Arcitopsia automatically applies additional compliance automation, audit depth, data residency enforcement, and multi-level approval chain logic for Enterprise, Multinational, and Mission-Critical customers — these run inside every job for those account types, adding real compute and LLM cost. Smaller deployments with simpler compliance profiles run leaner and cost less.
Base credit pricing is published in USD and applied globally. For multi-region deployments with specific data residency requirements, contact sales for a regional pricing assessment. Currency conversion for non-USD billing is handled at payment time using prevailing exchange rates.
BYOLLM customers pay reduced Intelligence Credit rates because the LLM inference cost component is removed from the credit price — you pay your LLM provider directly. Your credit balance and pack purchases work exactly the same way, just at the reduced BYOLLM rate. The 15% license fee discount is applied to your monthly subscription separately.
Yes. In your billing dashboard you can set a monthly credit spending cap, per-user spending limits, and low-balance alerts (email and in-app). You can also require admin approval for any job estimated above a credit threshold you define. This prevents unexpected spend while keeping the team productive within guardrails.
Your billing dashboard shows a complete credit ledger — every reservation, deduction, release, purchase, and monthly grant, with job-level detail including which workflow type, which user triggered it, and the actual vs. estimated credit cost. Monthly credit statements are also emailed to billing admins.
Free trials include a trial credit bundle valid for 30 days. Trial credits give you full access to all credit-bearing activities so you can evaluate Arcitopsia's AI outcomes. Trial credits cannot be topped up with purchased packs. If you convert to a paid plan before the trial ends, your first month's included credits activate and any remaining trial credits are replaced.
Book a demo to watch a compliance report or architecture document generated from a real registry — and see the credit estimate, reservation, and settlement flow live.
Free trial · No credit card required · Full credit access during trial
by iArchitron Software Inc. · iarchitron.ai