AI API cost calculator
Enter your monthly usage and see what every model would cost, from today's published prices.
Monthly cost by model
| # | Model | Per month |
|---|---|---|
| 1 | Grok Speech to Text
xAI · from $0.0017 / minute |
$5.00 |
| 2 | GPT-4o Mini Transcribe
OpenAI · $0.003 / minute |
$9.00 |
| 3 | Scribe v2
ElevenLabs · $0.0037 / minute |
$11.00 |
| 4 | GPT-Transcribe
OpenAI · $0.0045 / minute |
$13.50 |
| 5 | Gemini 3.5 Transcribe
Google · $0.005 / minute |
$15.00 |
| 6 | GPT-4o Transcribe
OpenAI · $0.006 / minute |
$18.00 |
| 7 | GPT-4o Transcribe Diarize
OpenAI · $0.006 / minute |
$18.00 |
| 8 | Whisper
OpenAI · $0.006 / minute |
$18.00 |
| 9 | Scribe v2 Realtime
ElevenLabs · $0.0065 / minute |
$19.50 |
| 10 | Gemini 3.5 Transcribe Live
Google · $0.009 / minute |
$27.00 |
| 11 | GPT-Live-Transcribe
OpenAI · $0.017 / minute |
$51.00 |
| 12 | GPT-Realtime-Whisper
OpenAI · $0.017 / minute |
$51.00 |
Frequently asked questions
How is the monthly cost calculated?
Input tokens are multiplied by the input price, output tokens by the output price, both per million tokens. The cached share of input uses the cached-input price when the provider publishes one, and the batch option uses batch prices where they exist. Prices are the ones valid today.
How many tokens is a word?
In English, a token is roughly three quarters of a word, so a million tokens is about 750,000 words. French and especially Arabic text usually takes more tokens for the same content, and each provider's tokenizer differs.
What is cached input?
When the start of a prompt repeats across requests, such as a long system prompt or a document, providers can reuse it and bill it at a much lower rate. Agents and document Q&A often reach 50 to 90% cached input.
Are free tiers and discounts included?
No. The calculator uses list prices in US dollars, before taxes, free allowances, committed-use discounts or regional surcharges.