LLM Cost Optimizer
Start trial

Cost Savings Calculator

Calculate potential savings with LLM Cost Optimizer based on your actual usage. All calculations use real API pricing from OpenAI, Anthropic, and Google.

How It Works

1

Enter your monthly API call volume

How many requests do you make per month?

2

Set average token counts

Typical input and output tokens per request

3

Choose your quality requirement

Low, medium, or high quality needs

4

See realistic cost comparison

Before vs. after with detailed breakdown

The calculator shows what you'd pay using only GPT-4 vs. using LLM Cost Optimizer's automatic model selection.

Calculate Your Savings

Total requests per month

Tokens sent to model

Tokens returned from model

Determines model selection

❌ Without LLM Cost Optimizer

£2700.00

per month (GPT-4 only)

• 100,000 calls/month
• 100% GPT-4 usage
• $0.03/1K input
• $0.06/1K output

✅ With LLM Cost Optimizer

£74.35

per month (optimized)

• 60% gemini 2.5 flash
• 30% claude 3.5 haiku
• 10% gpt 4o

💰 Your Monthly Savings

£2625.65
(97.2% reduction)

Cost Breakdown by Model

GEMINI 2.5 FLASH
60,000 calls (60%)
£5.85
CLAUDE 3.5 HAIKU
30,000 calls (30%)
£36.00
GPT 4O
10,000 calls (10%)
£32.50
Individual (£119)
0.05 months
to pay back
Pro (£179)
0.07 months
to pay back
Team (£3,000)
1.14 months
to pay back
Business (£12,000)
4.57 months
to pay back
* Calculations based on real API pricing as of November 2024. Your actual savings may vary based on usage patterns and quality requirements. Model selection percentages represent typical distribution patterns.
🔒

Included: Document Anonymization

Every license includes regex-based PII detection for protecting sensitive data:

Detection Capabilities

  • ✓Email addresses (pattern matching)
  • ✓Phone numbers (multiple formats)
  • ✓IPv4 addresses
  • ✓Numeric sequences (6+ digits)
  • ✓Custom regex patterns
  • ✓Reversible mapping for data restoration

How It Works

Local Processing
All PII detection runs on your machine using Python regex. No external services or cloud dependencies.
Three Integration Modes
CLI tool, Python library, or automatic API interception. Choose what fits your workflow.
✓ Included with all licenses
No per-document fees. No usage limits.
# Anonymize document before LLM call
$ llm-cost-opt anonymize ./document.txt --out ./redacted.txt

# Process with optimized LLM
$ llm-cost-opt optimize --prompt "Analyze: $(cat ./redacted.txt)"

✓ PII protected (local anonymization)
✓ Cost optimized (cheapest model selected)
✓ No sensitive data sent to external APIs

Note: Regex patterns may not catch all PII in all contexts. Review anonymized output for your specific use case. For compliance requirements, consult with legal counsel.

Try Live Demo →

API Pricing Reference

All models with current pricing (sortable table - click column headers):

Model ⇅Provider ⇅Input ($/1K) ↑Output ($/1K) ⇅Quality ⇅Best For
Gemini 2.5 FlashGoogle$0.000075$0.00030080%Lowest cost, high volume
GPT-4o-miniOpenAI$0.000150$0.00060070%Simple tasks, high volume
Claude 3.5 HaikuAnthropic$0.000800$0.00400080%Fast, cost-effective
Gemini 1.5 ProGoogle$0.001250$0.00500095%Long context, quality
GPT-4oOpenAI$0.002500$0.01000095%Balanced quality & cost
Claude 3.5 SonnetAnthropic$0.003000$0.015000100%Complex reasoning, analysis
GPT-4OpenAI$0.030000$0.060000100%Complex analysis, highest quality

Last Updated: November 23, 2024

Sources: All pricing verified directly from provider websites. Click provider names for official pricing pages.

Quality Score: Relative performance based on benchmarks. 100% = highest tier models.

Pricing Sources

All prices updated as of November 23, 2024:

Prices are subject to change. Calculator updates monthly.

⚠️ Important Disclaimer

These are estimates based on typical usage patterns. Your actual savings will vary based on:

  • Your specific prompts and responses
  • Quality requirements per task
  • Token usage patterns
  • Provider pricing changes

Use this as a guide, not a guarantee. We recommend testing with your actual workload.