API pricing calculator
API pricing calculator for AI apps
Estimate API cost before you ship. Compare AI API pricing, cost per request, cost per 1,000 calls, and monthly API spend across 111+ models from 8 providers, including OpenAI, Claude, Gemini, DeepSeek, Mistral, Grok, Meta, and Cohere.
Pricing data last updated September 22, 2026
API price
The provider rate, usually shown as dollars per 1M input tokens and 1M output tokens.
API cost per request
Your prompt and response token mix converted into a single request cost.
Monthly API cost
Request cost multiplied by real traffic, including daily calls, caching, and model routing.
How to calculate AI API pricing
- 1. Estimate input tokens. Count the user prompt, system prompt, retrieved context, and any tool results sent to the model.
- 2. Estimate output tokens. Forecast the response length for each request type. Output tokens often cost more than input tokens.
- 3. Apply model rates. Use the model input and output prices to calculate cost per request, then multiply by traffic.
- 4. Compare cheaper paths. Test budget models, prompt caching, batch APIs, and routing easy requests to efficient models.
Cheapest API pricing examples
Starting rates matter most for high-volume classification, extraction, routing, and background automation jobs.
Flagship API price comparison
Use flagship models when quality matters, then compare against efficient models before scaling traffic.
| Model | Input | Output | Context |
|---|---|---|---|
| Llama 4 Maverick | $0.27/1M | $0.85/1M | 1,000,000 tokens |
| Mistral Large 3 | $0.50/1M | $1.50/1M | 256,000 tokens |
| DeepSeek V4 Pro | $0.66/1M | $1.98/1M | 1,000,000 tokens |
| Gemini 3.8 Flash | $0.75/1M | $3.75/1M | 1,048,576 tokens |
| Gemini 3.7 Flash | $0.75/1M | $3.75/1M | 1,000,000 tokens |