Overview
AI Credits are the usage currency in Kodexa that track consumption of AI-powered features. Credits are consumed when processing documents with AI assistants, extracting data, running formulas with AI generation, and using other AI-powered capabilities.
What are AI Credits?
AI Credits represent computational resources used by Kodexa's AI features:
Document Processing - Credits consumed when AI extracts data from documents
AI Generation - Credits used for AI-suggested prompts and formulas
Document Chat - Credits consumed when querying documents via chat
Model Training - Credits used when AI learns from corrections
Credit consumption varies based on:
Document complexity (pages, content density)
Number of data elements extracted
AI model used (different models have different costs)
Task complexity (simple extraction vs. complex reasoning)
Credit Allocation
Initial Allocation
When you subscribe to Kodexa:
Credits are allocated to your Organization
Allocation amount depends on your subscription tier
Credits are shared across all users in your Organization
All Projects within the Organization share the credit pool
Monthly Replenishment
Credits refresh automatically:
Refreshed monthly based on your subscription
Unused credits roll over to the next month
No expiration on rolled-over credits while subscription is active
Credits accumulate if not fully used each month
Viewing Credit Balance
To check your current credit balance:
Navigate to your Organization homepage
Credit balance is displayed in the organization overview
Shows total available credits
Updated in real-time as credits are consumed
Adding Credits (Top-Up)
Organization Administrators can add credits at any time:
Steps to Add Credits
Navigate to your Organization settings
Click the Add Credits button
The Add Credits dialog opens
Review your current credit balance
Enter the number of credits to add
Click Add to complete the transaction
Add Credits Dialog
The dialog displays:
Organization Name - Confirms which organization receives credits
Slug - Organization identifier
Credits Input - Enter amount to add
Current Credit Balance - Shows current available credits
Credit History - Grid showing past credit transactions
Credit History
The transaction history shows:
Transaction Date - When credits were added
Credit Amount - Number of credits in the transaction
Sorted by date (newest first)
Tracks all credit additions
Credit Consumption
When Credits are Used
Credits are consumed when:
Processing Documents - Each document processed consumes credits based on complexity
AI Extraction - Using AI to extract data from documents
Formula Generation - AI suggests formulas or expressions
Semantic Definition Generation - AI generates prompts for data elements
Document Chat - Querying documents via the chat interface
Review AI Assistance - AI helps during data review workflows
Credit Usage Factors
Document Complexity
Single-page invoices: Lower credit consumption
Multi-page contracts: Higher credit consumption
Dense tables: More credits per extraction
Handwritten or poor quality: May require more credits
Data Definition Complexity
Few simple fields: Lower credit usage
Many complex fields: Higher credit usage
Nested groups: Additional credits
Repeating groups: Credits scale with instances
AI Model Selection
Different models have different credit costs
More powerful models consume more credits
Specialized models may have premium pricing
Managing Credit Usage
Monitoring Consumption
Check credit balance regularly in Organization homepage
Review which Projects consume most credits
Track consumption patterns over time
Set up alerts for low credit balances (if available)
Optimizing Credit Usage
Efficient Data Definitions
Extract only necessary data elements
Use formulas for calculated fields (no AI needed)
Consolidate similar elements
Remove unused or redundant fields
Smart Processing
Process documents in batches when possible
Use appropriate AI models for the task
Leverage templates for common document types
Review and correct efficiently to improve AI learning
Quality Over Quantity
Test with small batches first
Refine Data Definitions before large-scale processing
Use high-quality document scans to reduce retries
Train AI with corrections to improve efficiency
Credit Management Best Practices
Plan ahead - Estimate credit needs for upcoming projects
Monitor regularly - Check balance before starting large processing jobs
Top up proactively - Add credits before running out
Optimize definitions - Well-designed Data Definitions use fewer credits
Review consumption - Analyze which features consume most credits
Leverage rollover - Unused credits accumulate for future use
Batch processing - Process documents in batches for efficiency
Understanding Credit Transactions
Transaction Types
Subscription Allocation - Monthly automatic credit addition
Manual Top-Up - Credits added via Add Credits dialog
Consumption - Credits used during AI operations (tracked separately)
Transaction Details
Each credit transaction records:
Date and time of transaction
Number of credits added
User who initiated the transaction
Organization receiving the credits
Who Can Manage Credits
Organization Administrators
Only Organization Admins can:
View complete credit history
Add credits to the organization
Review credit consumption patterns
Manage credit allocation across projects
All Users
All organization members can:
View current credit balance
See credit usage in their projects
Process documents (consuming credits)
Request credit top-ups from admins
Credit Sharing Across Projects
Credits are shared at the Organization level:
All Projects draw from the same credit pool
No per-Project credit allocation
Encourages efficient credit usage across the organization
Admins can monitor which Projects consume most credits
What Happens When Credits Run Out?
If your credit balance reaches zero:
AI-powered processing stops
Document upload continues (documents stored, not processed)
Existing extracted data remains accessible
Formulas and manual entry continue to work
Admin must add credits to resume AI processing
No data loss occurs
Tips
Credit balance updates in real-time as you process documents
Unused monthly credits roll over - they don't expire
Organization Admins can add credits at any time
Credit consumption varies by document and task complexity
Well-designed Data Definitions are more credit-efficient
Test with small batches to estimate credit needs for large jobs
Review credit history to understand usage patterns
Consider adding credits proactively for large processing projects
AI learning from corrections improves accuracy and may reduce future credit usage