Free · Open · Updated monthly

Calculate AI API costs in seconds.

Compare pricing across 10+ models from OpenAI, Anthropic, Google, DeepSeek, xAI and Mistral — including cached input and Batch API discounts.

  • Exact tokenization for OpenAI models, transparent estimates for others
  • Side-by-side caching & batch savings
  • Monthly cost forecasts for high-volume workloads
Updated 2026-05-12·USD / CNY / EUR / GBP / INR
✓ Exact tokenizer·$5.00 in·$30.00 out (per 1M)
Quick start with a use case
Total cost per call$0.0200
Input$0.005000
Output$0.0150
Cost comparison
Standard
$0.0200
With Caching
$0.0177
Save 11% ↓
With Batch
$0.0100
Save 50% ↓
Side by side

All 10 models, ranked by cost

Plug in your token usage. Sort, filter, hide. The cheapest pick for your workload is at the top.

All models compared

Sorted by cost for your token usage. Click column headers to re-sort.

Filter by provider
DeepSeek V3.2
DeepSeek · balanced
$0.2800$0.4000$0.000480
GPT-5 mini
OpenAI · small
$0.2500$2.0000$0.001250
Mistral Large 3
Mistral · balanced
$0.5000$1.5000$0.001250
Gemini 3 Flash
Google · balanced
$0.5000$3.0000$0.002000
o4-mini
OpenAI · reasoning
$1.1000$4.4000$0.003300
Claude Haiku 4.5
Anthropic · small
$1.0000$5.0000$0.003500
Gemini 3.1 Pro
Google · flagship
$2.0000$12.0000$0.008000
Grok 4
xAI · flagship
$3.0000$15.0000$0.0105
Claude Opus 4.7
Anthropic · flagship
$5.0000$25.0000$0.0175
GPT-5.5
OpenAI · flagship
$5.0000$30.0000$0.0200
Plan ahead

Forecast your monthly bill

Enter your daily call volume and average tokens. See the monthly damage across up to 5 models with savings highlighted.

Monthly cost forecast

Estimate your monthly bill across up to 5 models.

3/5 selected
Opus 4.7Anthropiccheapest
$300.00/mo
per call: $0.0100year: $3650.00
GPT-5.5OpenAI
$345.00/mo
per call: $0.0115year: $4197.50
Choosing Opus 4.7 over GPT-5.5 saves $45.0000/month ($540.00/year).
Last verified 2026-05-12Schema v2.0Updated monthlyOpen source on GitHub