Billing & Credits

Aexol uses a credit-based billing system. Every AI model operation consumes credits based on the number and type of tokens processed. This page explains how credits work, how they're calculated, and what each model costs.

How Credits Work

Every time you run an AI operation — code generation, refinement, agent interaction — the system tracks the tokens consumed and converts them into credits using per-model rates.

Credit Calculation Formula

Credits = (inputTokens / 1,000,000) × creditInputPer1M

(outputTokens / 1,000,000) × creditOutputPer1M
(cacheWriteTokens / 1,000,000) × creditCacheWritePer1M
(cacheReadTokens / 1,000,000) × creditCacheReadPer1M

Credits are rounded to 2 decimal places. If a model has no credit rates configured, no credits are deducted.

Token Types

Note: Cache write/read rates only apply to providers that support prompt caching (e.g., Anthropic). If a model has no cache rates set, cached tokens are billed at the standard input rate.

Live Model Credit Rates

The table below shows credit costs for every available model. Values are credits per 1,000,000 tokens. Rates are fetched in real-time from the public API.

Reading the Credit Table

• Input — Credits for prompt tokens (your spec, context, knowledge base)
• Output — Credits for generated tokens (the AI's response)
• Cache Write — Credits for filling the cache (first request in a session)
• Cache Read — Credits for reading from the cache (subsequent requests)

Models shown with reduced opacity have no credit rates configured — using them won't consume credits (reserved for future pricing).

Plan Limits

Free Plan

• 500 credits per month
• Operations are blocked once you reach the limit
• Ideal for evaluation and small projects

Pro Plan

• 10,000 credits per month
• Overage is billed via Stripe (usage-based)
• Priority support
• Team collaboration features
• Multiple projects and knowledge base documents

Tracking Your Usage

In Studio

Go to Studio → Billing to see your current usage:

• Credits used / credits remaining
• Token breakdown (input, output, cache)
• Percentage of monthly limit consumed
• Billing period dates

Team Owners

Team owners can manage subscriptions at Studio → Teams → Billing:

• View current plan and usage
• Upgrade to Pro
• Cancel or resume subscriptions
• Access Stripe billing portal

Usage Per Inference Task

Each inference task shows its exact credit consumption. View it:

1. Go to Studio → Tasks
2. Click on any completed task
3. See creditsUsed in the task details panel

The creditsUsed field is part of the TokenUsage type in the GraphQL API.

API Access

The backend exposes public queries for billing information:

Available Plans

query {
  availablePlans {
    id
    name
    monthlyPrice
    creditsIncluded
    tokensIncluded
    features
  }
}

Team Subscription (authenticated)

query($teamId: ID!) {
  teamSubscription(teamId: $teamId) {
    plan
    status
    usage {
      creditsUsed
      creditsLimit
      percentUsed
      totalTokensUsed
    }
  }
}

Available Models with Credit Rates (public)

query {
  availableBaseModels {
    name
    provider
    creditInputPer1M
    creditOutputPer1M
    creditCachedInputPer1M
    creditCacheReadPer1M
    creditCacheWritePer1M
  }
}

Next Steps

• Code Generation — Start generating code
• Teams & Collaboration — Set up team billing
• Artifacts — Review generated output