AI Model Cost Calculator

Calculate AI model API costs for your exact usage. No sign-up, no tracking — just accurate math.

New to AI? Pick a Quick Preset. Have an account? Check your provider dashboard for real usage numbers.

Tokens sent in your API requests
Tokens generated by the model
Number of API calls per month

Quick Presets

Select Models

30 models · Updated 2026-07-044 / 5
Anthropic5 models
OpenAI7 models
Google3 models
DeepSeek2 models
Alibaba2 models
Moonshot AI1 models
MiniMax1 models
Zhipu AI1 models
Mistral2 models
Meta2 models
Cohere1 models
Amazon1 models
SpaceXAI2 models

Prompt Caching

Drag to estimate your cache hit rate.

0%

0% = every request is unique. 50%+ with fixed system prompts. 90%+ for coding assistants.

Your Estimated Costs

4 models selected · Monthly estimate

Claude Sonnet 4.6

Anthropic

Estimated Monthly Cost

$45.00

Input cost$15.00
Output cost$30.00

GPT-5.4

OpenAI

Estimated Monthly Cost

$42.50

Input cost$12.50
Output cost$30.00

Gemini 3.1 Pro

Google

Estimated Monthly Cost

$34.00

Input cost$10.00
Output cost$24.00

DeepSeek V4 Pro

DeepSeek

Estimated Monthly Cost

$3.92

Input cost$2.17
Output cost$1.74

💱 Costs shown in USD. Convert to your currency →

Why Compare AI Model Pricing?

AI API pricing is fragmented. Each provider charges differently for input tokens, output tokens, prompt caching, batch processing, and long-context usage. A model that looks cheap at first glance can cost 2–100× more than an alternative for your specific usage pattern.

This calculator compares 30 models from 13 providers — including Anthropic (Claude), OpenAI (GPT-5), Google (Gemini), DeepSeek, xAI (Grok), Meta (Llama), Mistral, Alibaba (Qwen), Cohere, Amazon (Nova), Moonshot AI, and MiniMax. Every price is sourced directly from official API pricing pages and tracked in our public changelog.

Popular Model Comparisons

Claude Sonnet 4.6 vs GPT-5.4

Flagship face-off →

Claude Fable 5 vs GPT-5.6 Sol

Most capable models →

Claude Opus 4.8 vs Claude Sonnet 4.6

Anthropic tier comparison →

Claude Sonnet 4.6 vs Gemini 3.1 Pro

Cross-provider comparison →

Claude Haiku 4.5 vs GPT-5.6 Luna

Best value models →

Claude Opus 4.8 vs GPT-5.4

Premium showdown →

Browse all 435 model comparisons →

How to Use

  1. Enter your usage — input tokens, output tokens, and monthly requests. Not sure? Pick one of 8 presets.
  2. Select models — compare up to 5 at once. Costs update instantly.
  3. Toggle advanced options — factor in prompt caching, batch processing, or long-context pricing.
  4. Dig deeper — click into any model pair for full specs, benchmarks, and provider details.

What Makes This Different

  • Not just a table — plug in your own token counts, not generic estimates.
  • All pricing dimensions — input, output, cache read/write, batch, and long-context surcharges.
  • 435 comparison pages — every model pair has a dedicated page with full cost breakdowns.
  • Prices verified manually — we check official provider pricing pages, not scrapers.
  • No signup, no tracking — just math. No account needed.