Pricing @ Arpia Platform

Arpia Credits

At ARPIA we believe in adding value into your organization, for this we have designed a credit consumption system that is clear and focused on pricing based on value generation.

Credits are consumed based on items and value units. The following page explains these items and their credit costs. The groups of items are:

GroupDescription
Value Units (VUs)Arpia Platform Value Units are based on features used in your workarea(s).
Credit UsageCredit value for VUs, Hours or Unit Of Measure for resources used in your workarea(s).
Core LLMsArpia LLM Core Models available for users to start using LLM models into their projects.
StorageThe storage capacity by workarea assigned to the customer.

Credits are the currency used for usage-based billing in ARPIA Data Cloud.

👍

Credit Hard Limit

Hard limits are available per workarea so you can set a VUs hard limit and the platform will limit the creation of more VU items once the hard limit is reached.


Arpia Value Units (VUs)

If your account is set to be billed with VUs (Value Units) then this table applies. Value Units is the default way for billing our platform value into your organization in a transparent and balanced method on charging only if value is delivered by our Platform. You may analyze in the workarea usage page information on how Value is being generated, where and costs for each area.

Arpia Value Units is an innovative way of measuring the value that the platform generates into the organization in a time frame. It's calculated by the resources and features used in workareas. The following table shows the VU price for each item.

ItemDescriptionVUs
Work Area UsersAverage count of users that belong to the workarea. Invited users charged in other workareas are not billed to avoid billing duplication.20.00
AI Worker / AI OperatorAverage count of active workshop projects that have AI Workers or AI Operators in the workarea. Multiple Workers / Operators Objects in one workshop project count as one.200.00
Workshop ProjectsAverage count of active workshop projects in the workarea.50.00

For DataApps the Value Units are assigned by type of screens or widgets used in the application, based on the following table:

ItemDescriptionVUs
AI AssistantAverage count of AI Assistant agents used in the workarea DataApps. The amount of Workflow steps does not affect the cost of this item.50.00
DashboardsAverage count of dashboard screens used in the workarea DataApps. These dashboard screens can have x amount of dashlets, charts, pivot tables etc. These sub items do not count on VUs.15.00
Page LayoutAverage count of page layout screens used in the workarea DataApps. These page layout screens can have x amount of sub items that does not affect the value of the item.10.00
Custom CodeAverage count of custom code screens in the workarea DataApps.15.00
Calendar ScreenAverage count of calendar screens in the workarea DataApps.10.00
Map ScreenAverage count of map screens in the workarea DataApps.100.00
Data EditAverage count of data edit tables used in the workarea DataApps.5.00
Profile SecurityAverage count of this feature used in the workarea DataApps.50.00
Public AppAverage count of this feature used in the workarea DataApps.100.00

If there is a feature not listed in this table its cost is 0 VUs.

👍

VUs Billing Process & Hard limits

The billing process of VUs is as follows: every day the platform generates usage of each feature and saves the VUs used for that day. For the credit consumption process the system calculates the average usage within the time frame (month) of VUs and applies credit costs. This way if you decrease or increase the features used your billing will be reflected since the day you apply the changes.


VUs Credit Price

All resources used to generate value into your organization using Arpia Platform have a value in credits. Your workarea will be charged by usage of these resources per month.

Resource TypeBilling CycleQuantityUoMCost (Credits)
Value Units (VUs)Monthly1VU1

VUs usage and billing information is well documented in your workarea usage page. Hard limits can be set for limiting VUs overflow for controlling fixed billing costs.


Computing Hours @ Arpia Cloud

If your plan is billed by Usage Computing Credits then the following conversion table outlines the costs in credits for computing usage when applicable.

Resource TypeBilling CycleQuantityUoMCost (Credits)
Computing Hours does not apply when VUs are Used.Monthly Charge1Hour0.90

Computing hours are calculated by usage in the platform and its components. Usage credits dashboard available for clear insights.

❗️

Computing Hour does not apply if your account is billed by Value Units. Computing credits can't be controlled and are best for scenarios where customers prefer this model rather than fixed Value Units Billing.


GPU, Storage, AI LLM & Other Components

The following section describes the pricing and credit costs for components that are part of the offering of Arpia Platform Cloud Services.

GPU Droplets

When you need to use GPU and other special components for your organization value objectives, you can add them through customer support chat or account executive and these items will be charged as follows.

Resource TypeBilling CycleQuantityUoMCost (Credits)
AI Dedicated GPU Hours (AI/ML Models, Local GPU)Usage1HourGPU Type Based

LLMs Core Models @ Arpia Cloud

Arpia provides core LLM models from most providers. If a customer configures its own workarea models then no tokens are charged and usage is limited to the computing hours. If AI is used and no token is provided, Arpia uses its own global shared tokens for user simplicity and they are billed by token usage.

📊 Model Classification for Enterprise

For Enterprise customers, we have organized LLM models into two distinct categories to provide clarity on what's included in your plan versus additional consumption.


🟢 ARPIA Base LLM Models

Models included in Enterprise plans

These models are included in the monthly token allocation according to your Enterprise plan. They do not generate additional charges as long as you stay within your contracted token limit.

ModelTokensCost (Credits/1M)Features
OpenAI
openai/gpt-4o-mini1M0.75Compact general-purpose model
openai/o1-mini1M5Basic reasoning
openai/o3-mini1M5Latest generation compact
openai/gpt-4o1M5Standard multimodal model
openai/gpt-5-nano1M5Ultra-efficient version
Anthropic
anthropic/claude-haiku-4-51M5Fast and efficient responses
Cerebras
cerebras/gpt-oss-120b1M1High-performance open source model
cerebras/llama3.1-8b1M1Optimized compact model
cerebras/llama-3.3-70b1M1Medium-large model
Groq
groq/llama-3.3-70b-versatile1M1Versatile and fast
groq/llama-3.1-8b-instant1M0.75Instant responses
groq/gpt-oss-120b1M1Accelerated open source
Google
google/gemini-2.0-flash-0011M0.75Fast multimodal
google/gemini-2.5-flash-preview-05-201M0.75New generation flash
Mistral AI
mistralai/codestral-25011M0.75Code specialized
mistralai/mistral-nemo1M0.75Efficient compact model
mistralai/pixtral-large-24111M4Advanced multimodal
Grok
grok/grok-4-1-fast-reasoning1M0.75Fast reasoning
grok/grok-4-1-fast-non-reasoning1M0.75Standard processing

🔴 ARPIA Frontier LLM Models

Pay-per-use additional models 💳

These models represent the most advanced technology available and are NOT included in the base Enterprise token package. They are billed independently based on actual consumption, adding to your monthly invoice.

ModelTokensCost (Credits/1M)Billing TypeFeatures
OpenAI
openai/gpt-5-mini1M10🔴 Per ConsumptionEfficient GPT-5 model
openai/gpt-51M25🔴 Per ConsumptionGPT-5 flagship model
openai/gpt-5.21M25🔴 Per ConsumptionLatest GPT-5 version
openai/o11M37.5🔴 Per ConsumptionMaximum advanced reasoning
Anthropic
anthropic/claude-sonnet-4-51M17.5🔴 Per ConsumptionBalanced capability-speed
anthropic/claude-opus-4-51M25🔴 Per ConsumptionMaximum analytical capability
Google
google/gemini-2.5-pro-preview1M10🔴 Per ConsumptionNew generation Pro
google/gemini-3-pro-preview1M12🔴 Per ConsumptionNext generation Gemini

📋 Enterprise Business Rules

✅ Base Models (Included)

  • Included in: All Enterprise plans according to contracted tokens
  • Billing: No additional charge within monthly token limit
  • Consumption: Counts against plan token limit
  • Ideal for: Daily operations, automations, basic assistants

💰 Frontier Models (Per Consumption)

  • Included in: No plan (always pay-per-use)
  • Billing: Additional charges based on tokens consumed
  • Consumption: Does NOT count against base plan token limit
  • Invoice: Separate line item on monthly invoice: "LLM Frontier - Additional Consumption"
  • Ideal for: Complex analysis, advanced reasoning, specialized use cases

🎯 Usage Recommendations

To maximize the value of your Enterprise plan:

  1. Use Base models for 80-90% of your daily operations
  2. Reserve Frontier models for cases requiring exceptional capabilities
  3. Monitor consumption in the workarea usage dashboard
  4. Configure alerts for Frontier model consumption

📘

BYM (Bring Your Model)

When users add their own models to Arpia Work area there will be no charge done for LLM usage using that BYM Model. Only Computing Hours or VUs will be billed as applicable.


Storage @ Arpia Cloud

The following conversion table outlines the costs in credits for storage usage when applicable. We have two options for storage: Shared and Dedicated.

Shared Storage

This storage is multi-tenant for organizations that are not on a Tier plan. It's fully isolated but the resources are shared between tenants.

Resource TypeBilling CycleQuantityUoMCost (Credits)
Data Storage (Shared/Dedicated SingleStore Cluster)Monthly Daily Avg. Charge10GB5.00
Data Storage (Shared/Dedicated MySQL/MariaDB Cluster)Monthly Daily Avg. Charge10GB2.50
Object StorageMonthly Daily Avg. Charge10GB1.00

Dedicated Storage

On Tier plans our enterprise customers can opt for dedicated resources. Please use the following table for calculating your dedicated infrastructure. For HA setups please add +1 node for each existing node.

Resource TypeBilling CycleQuantityUoMCost (Credits)
Dedicated Storage Node (SingleStore Node 32GB 8vCPU)Monthly Charge1Node750
Dedicated Storage Node (MySQL/MariaDB Single Node 32GB 8vCPU)Monthly Charge1Node250
Dedicated Object Storage Node 8GB 4vCPUMonthly Charge1Node150

The storage of each node will be charged using the pricing table for Shared Storage.

📘

For accounts in Tier Contracts you may be entitled to extra discounts in Computing and AI Computing Credits, please refer to your Account Manager.


Credit Billing Policy & Terms

To ensure clarity in billing, the following policies apply:

  1. Credit Rollover:
    • Annual plans only. If you consume fewer credits than purchased, they roll over to the next billing period. Monthly plans do not allow rollovers.
  2. Credit Refunds:
    • There are no refunds as credits are billed by usage of a past period. Adjustments may apply if consumption was generated by system error.
  3. On the 5th of each month, the billing system generates and sends invoices of credits based on the previous month's usage.
  4. Customers can pay their invoices securely through credit card powered by Stripe.
  5. Customers that have more than 2 invoices overdue can experience service interruption.
  6. If Customers have a prepaid amount it will be added to the customer account and recurring usage billing will be deducted from this prepaid balance.

Account Billing & Usage Support

For assistance, contact billing support via the Support Chat on the ARPIA Web Platform, accessible through the main menu.