Skip to main content

Billing Guide

Transparent pricing with detailed cost attribution Understand how costs are calculated and billed.

Billing Model​

Korad.ai uses a transparent billing model with three components:

  1. Actual Cost - What we pay to LLM providers
  2. Profit Margin - Our service fee (default 1.5x)
  3. Your Bill - Actual cost × profit margin

Cost Attribution​

Every response includes detailed cost headers:

X-Korad-Original-Tokens: 150006
X-Korad-Optimized-Tokens: 32743
X-Korad-Tokens-Saved: 117263
X-Korad-Reduction-Pct: 78%
X-Korad-Savings-USD: $0.029214
X-Korad-Theoretical-Cost: $0.037502 # What you would have paid
X-Korad-Actual-Cost: $0.008288 # What we actually paid
X-Korad-Billed-Amount: $0.056253 # What we charge you
X-Korad-Profit-Margin: 1.50x

Pricing Formula​

theoretical_cost = (input_tokens / 1,000,000) × input_price
+ (output_tokens / 1,000,000) × output_price

actual_cost = theoretical_cost × (1 - optimization_savings)

billed_amount = actual_cost × profit_margin

Example Calculation​

Request: 150,006 input tokens, using Claude Sonnet ($3.00/1M input)

Without optimization:

theoretical_cost = (150,006 / 1,000,000) × $3.00 = $0.450
billed_amount = $0.450 × 1.5 = $0.675

With 78% optimization:

actual_cost = $0.450 × (1 - 0.78) = $0.099
billed_amount = $0.099 × 1.5 = $0.149

Your savings: $0.675 - $0.149 = $0.526 (78% savings)

Provider Pricing​

Anthropic​

ModelInputOutput
Claude Opus$15.00/1M$75.00/1M
Claude Sonnet$3.00/1M$15.00/1M
Claude Haiku$0.25/1M$1.25/1M

OpenAI​

ModelInputOutput
GPT-4o$2.50/1M$10.00/1M
GPT-4o Mini$0.15/1M$0.60/1M

DeepSeek​

ModelInputOutput
DeepSeek Chat$0.14/1M$0.28/1M
DeepSeek Reasoner$0.55/1M$2.19/1M

Google Gemini​

ModelInputOutput
Gemini 2.5 Pro$1.25/1M$10.00/1M
Gemini 2.5 Flash$0.075/1M$0.30/1M

Optimization Savings​

Tier 2: Vanishing Context (99%)​

Original: 150,000 tokens → Optimized: 1,500 tokens

theoretical_cost = (150,000 / 1,000,000) × $3.00 = $0.450
actual_cost = (1,500 / 1,000,000) × $3.00 = $0.0045
billed_amount = $0.0045 × 1.5 = $0.00675
savings = $0.450 - $0.00675 = $0.44325 (99%)

Tier 3: Recursive RLM (97%)​

Original: 200,000 tokens → Optimized: 6,000 tokens

theoretical_cost = (200,000 / 1,000,000) × $3.00 = $0.600
actual_cost = (6,000 / 1,000,000) × $3.00 = $0.018
billed_amount = $0.018 × 1.5 = $0.027
savings = $0.600 - $0.027 = $0.573 (97%)

Tier 5: Savings Slider (89%)​

Original: 150,000 tokens → Optimized: 16,384 tokens (extreme)

theoretical_cost = (150,000 / 1,000,000) × $3.00 = $0.450
actual_cost = (16,384 / 1,000,000) × $3.00 = $0.049
billed_amount = $0.049 × 1.5 = $0.074
savings = $0.450 - $0.074 = $0.376 (89%)

Billing Cycles​

Monthly Billing​

  • Billing Period: Calendar month
  • Invoice Date: 1st of each month
  • Payment Due: 15 days after invoice

Minimum Commitment​

Some plans have monthly minimums:

PlanMinimumIncludes
Starter$0Pay-as-you-go
Pro$50/month100K included tokens
Enterprise$500/month1M included tokens

Payment Methods​

  • Credit Card (Visa, Mastercard, American Express)
  • ACH Transfer (Enterprise)
  • Wire Transfer (Enterprise)

Invoices​

Each invoice includes:

  • Summary: Total cost, taxes, credits
  • Usage Breakdown: By model, by date
  • Optimization Report: Total savings achieved
  • Transaction Log: Detailed request history

Sample Invoice​

KORAD.AI - Invoice #2025-02-001
Billing Period: February 2025
─────────────────────────────────────────

USAGE SUMMARY
─────────────────────────────────────────
Model Tokens Cost
─────────────────────────────────────────
claude-opus-4-20250514 1.2M $18.00
claude-sonnet-4-5-20250929 5.5M $16.50
claude-haiku-4-5-20251001 8.3M $2.08
gpt-4o 2.1M $5.25
─────────────────────────────────────────
Subtotal $41.83
Optimization Applied -$26.45
Actual Cost $15.38
Profit Margin (1.5x) $23.07
─────────────────────────────────────────
TOTAL DUE $38.45

OPTIMIZATION SAVINGS
─────────────────────────────────────────
Original Cost: $41.83
Optimized Cost: $15.38
Your Savings: $26.45 (63%)
─────────────────────────────────────────

Cost Monitoring​

Real-Time Monitoring​

curl http://localhost:8081/api/billing/usage

Response:

{
"current_month": {
"total_cost": 38.45,
"total_savings": 26.45,
"total_requests": 15420,
"breakdown_by_model": {
"claude-opus-4-20250514": 18.00,
"claude-sonnet-4-5-20250929": 16.50
}
}
}

Budget Alerts​

Set up alerts at specific thresholds:

  • 50% of budget
  • 80% of budget
  • 100% of budget (hard limit)

Refund Policy​

  • Service Credits: Issued for downtime
  • Prorated Refunds: For mid-month cancellations
  • Overage Refunds: If billed in error

Support​

For billing questions:


Transparent pricing with detailed cost breakdown on every request.