About This Tool

This calculator helps developers, product managers, and teams estimate their AI API costs before committing to a provider. Enter your expected usage pattern and compare costs across every major model side by side.

100% Client-Side

All calculations happen in your browser. No data is transmitted to any server.

How Costs Are Calculated

For each model, the calculator computes: (requests/day x avg_input_tokens / 1,000,000 x input_price) + (requests/day x avg_output_tokens / 1,000,000 x output_price). Daily cost is then multiplied by 30 for monthly and 365 for yearly estimates.

Pricing Tiers

• Standard — Real-time API pricing as published by each provider.
• Batch (50% off) — Most providers offer batch/async processing at half price. Turnaround is typically 24 hours.
• Cached input (90% off input) — Prompt caching stores your system prompt and charges ~10% of the standard input rate for subsequent requests. Output pricing remains unchanged.

Data Sources

Pricing is sourced from each provider's official API pricing page and cross-referenced with OpenRouter's passthrough rates. Last verified: September 2026. Meta/Llama models use representative pricing from AWS Bedrock and Together AI.

Limitations

• Does not account for reasoning token overhead (o-series, R1-series thinking tokens).
• Does not account for long-context surcharges (Gemini 200K+ premium).
• Batch and caching discounts are approximate — exact rates vary by provider.
• Does not include tool use, image, audio, or video token costs.

Contact

Questions or feedback? Use the feedback link in the navigation.