How to use the AI API Cost Calculator
1. Enter your expected input tokens per request (your prompt length) and output tokens per request (the response length).
2. Set the number of requests (per month, per day, or any period you want to compare).
3. The cost table updates instantly — sorted cheapest first by default.
Tip: 1 token ≈ 4 characters or ¾ of a word. A typical 1-page document ≈ 750 tokens.
What is it?
AI API pricing varies enormously across providers — the cheapest option can be 1000× less expensive than the priciest for the same task. Understanding cost before you scale is critical for production applications. This calculator uses the latest published list prices for OpenAI, Anthropic, Google, Meta (Groq/Together), Mistral, and Cohere.
Prices are per million tokens (input and output priced separately). Input tokens are your prompt; output tokens are the response generated by the model. Most applications have 3–10× more input tokens than output tokens for typical chat tasks.
Note: actual costs may differ due to caching, batching discounts, reserved pricing, or model versions. Always check the provider's official pricing page for the most up-to-date rates.
Example
Input tokens: 500 (prompt) Output tokens: 300 (response) Requests: 10,000/month
Gemini 1.5 Flash-8B: $0.02/month GPT-4o mini: $0.93/month GPT-4o: $15.50/month Claude 3 Opus: $97.50/month
Model choice is often the single biggest cost lever in production AI systems.
Benefits
Generate accurate code from your actual data
TypeScript interfaces, Python dataclasses, and OpenAI function schemas are generated directly from your real JSON — accurate field names, correct types, no guesswork.
Control LLM API costs before you scale
Token counts and per-call cost estimates for GPT-4o, Claude 3.5, and Gemini let you optimise prompt size and model choice before a feature goes to production.
Ready-to-run code and configs, not boilerplate
Every tool produces output you can paste directly into your project — API request code, prompt templates, chunk configurations — adapt rather than write from scratch.
Purpose-built for AI engineering workflows
From RAG chunking strategy to prompt template selection to function calling schema design — tools built for the specific problems AI engineers encounter daily.
Frequently asked questions
How do I estimate my token count?
A rough rule: 1 token ≈ 4 characters (English text). So a 1000-character prompt ≈ 250 tokens. A full page of text ≈ 500–750 tokens. For precise counts, use our Prompt Token Counter tool.
Why are input and output tokens priced differently?
Generating output tokens requires more computation than reading input tokens. Most providers charge 3–5× more per output token than per input token. For tasks that generate short responses (classification, extraction), input cost dominates. For long-form generation, output cost dominates.
What is context caching and does it affect cost?
Context caching (offered by Anthropic and Google) stores a prompt prefix server-side so you don't re-pay for it on every request. If your system prompt is large and constant, caching can cut costs by 80–90%. This calculator shows uncached list prices.
Are these prices current?
Prices are based on publicly available list prices. AI pricing changes frequently as models are updated and competition increases. Always verify on the official provider pricing pages: platform.openai.com/pricing, anthropic.com/pricing, ai.google.dev/pricing.
What is the difference between batch and realtime pricing?
OpenAI and Anthropic offer batch APIs at roughly 50% discount for non-realtime workloads (results delivered within 24 hours). This calculator shows realtime pricing. If latency isn't critical, batch APIs can halve your costs.