Free Tool

LLM Cost Calculator

Calculate the monthly cost of your LLM API usage across providers. Enter your usage and compare side-by-side.

Usage Configuration

tokens

Approx. Word Count:~3,750 words

tokens

Approx. Word Count:~1,500 words

calls

Total volume: 35,000,000 tokens

Target Models for Comparison

Anthropic

OpenAI

Google

Mistral AI

Groq

Together AI

DeepSeek

Monthly Billing Projections

Relative comparison sorted by cheapest API provider

⚡ Save up to$218.50/mo
1

Gemini 2.0 Flash

Google
$0.10/M in · $0.40/M out$6.50/mo
2

DeepSeek V3

DeepSeek
$0.27/M in · $1.10/M out$17.75/mo
3

GPT-4o

OpenAI
$2.50/M in · $10.00/M out$162.50/mo
4

Claude 3.5 Sonnet

Anthropic
$3.00/M in · $15.00/M out$225.00/mo

Prices reflect public standard endpoints as of May 2026. Rate structures like Anthropic cache writes or prompt caching are not included in estimates.

How LLM pricing works

Most LLM providers charge per million tokens — separately for input (your prompt) and output (the model's response). Output tokens are usually 3–5x more expensive than input tokens because generation is computationally heavier.

The formula is simple: 

monthly_cost = (input_tokens / 1M) * input_price + (output_tokens / 1M) * output_price

Multiply by your monthly request volume to get your total bill.

Key cost factors

  • Input vs output ratio: Prompts with short outputs (summarization, classification) are cheaper than open-ended generation
  • Model tier: Frontier models (Claude 3 Opus, GPT-4 Turbo) cost 10–100x more than small models (Haiku, GPT-4o Mini)
  • Prompt caching: Anthropic and OpenAI offer caching discounts of up to 90% on repeated context
  • Batch mode: Many providers offer 50% discounts for non-real-time batch workloads

Related tools

Related guides