How is AI API cost calculated?

AI API is billed per token, commonly per million tokens in USD. Monthly cost = (monthly input tokens / 1M × input price) + (monthly output tokens / 1M × output price). Models with prompt caching can further reduce costs by 50-90% on cached tokens.

Can I trust relay provider pricing?

TokenAPI Scan captures relay pricing from public rate pages with periodic cross-checks, but actual billing is at the relay's discretion — some providers may use different token counting methods. We recommend screening providers via the leaderboard first, then using the calculator to compare costs.

Why do prices for the same model vary 10x across relays?

Reasons include: 1) Upstream cost differences (official / tier-1 / tier-2 distribution); 2) Possible model downgrading or shared pool dilution; 3) Volume plans vs pay-as-you-go; 4) Whether retry and caching costs are included. Extremely low prices should be verified — our leaderboard and detection reports can help.

💰 AI API Cost Calculator

Enter your monthly token usage and compare GPT-5 / Claude 4 / Gemini 3 / DeepSeek across 0+ relay providers. Based on real-time public rate snapshots from TokenAPI Scan.

Monthly Input Tokens

1M 10M 100M 1B

Monthly Output Tokens

0.2M 2M 20M 200M

Monthly Cached Tokens (optional)

0 5M 50M

Add Model to Compare

claude-opus-4-7 claude-opus-4-6 claude-sonnet-4-6 claude-opus-4-8 claude-fable-5 claude-haiku-4-5-20251001

All calculations run locally in your browser, results update in real-time

Note on methodology: Prices in USD / 1M tokens, sourced from each relay provider's public rate page. Cached prices are not published by some providers; where missing, they are estimated from input price. Actual billing is subject to each provider's terms.

📊 Monthly Cost Overview

—Select models to calculate

🔍 Price Breakdown by Relay (sorted by total cost)

📚 Next Steps After Choosing a Relay

Check the relay's Red/Black List Report for authenticity, protocol support, and model coverage
Browse the Full Price Database — 1200+ models × 60+ relays
Read the Selection Guide to avoid low-price traps and shared pool dilution
Use AI Key Leak Detection to check if your existing keys are exposed

📬

Want to know when prices drop? Subscribe to price alerts — get notified the moment your models drop below your threshold. Free, unsubscribe anytime.

Subscribe Now →