AI Cost Calculator

📖 2 min read toolsreferencecost

Estimate your monthly AI API costs based on requests, tokens, and model choices - Claude, GPT, Gemini, and DeepSeek.

Key Takeaways

Use the sliders to estimate monthly costs for any combination of requests and tokens
DeepSeek V4 Flash and GPT-5.5 Instant are the cheapest options
Model routing can reduce costs by 60-80%

Estimate your monthly costs based on requests, tokens, and model choices.

Interactive Calculator

Usage Scenario

Daily Requests

100 5,000/day

Avg Input Tokens per Request

1,000 50,000

Avg Output Tokens per Request

500 10,000

Quick scenarios:

Monthly Cost

Anthropic

Claude Sonnet

$31.50

$3/$15 per 1M Default

Anthropic

Claude Haiku

$10.50

$1/$5 per 1M

Anthropic

Claude Opus

$52.50

$5/$25 per 1M

OpenAI

GPT-5.4 mini

$9.00

$0.75/$4.50 per 1M Fast

OpenAI

GPT-5.5

$60.00

$5/$30 per 1M Balanced

OpenAI

GPT-5.4

$30.00

$2.50/$15 per 1M Value

Google

Gemini 3.1 Pro

$24.00

$2/$12 per 1M 1M ctx

DeepSeek

DeepSeek Flash

$0.84

$0.14/$0.28 per 1M Cheapest

DeepSeek

DeepSeek V4 Pro

$2.61

$0.435/$0.87 per 1M Premium

xAI

Grok 4.3

$7.50

$1.25/$2.50 per 1M Real-time

Alibaba

Qwen 3.6

$3.92

~$0.33/$1.95 per 1M

Google

Gemini 3.5 Flash

$18.00

$1.50/$9 per 1M Fast

Moonshot AI

Kimi K2.6

$5.55

~$0.60/$2.50 per 1M 256K ctx

Cost Breakdown Formula

Daily cost = (requests × input_tokens × input_price/1M) + (requests × output_tokens × output_price/1M)
Monthly cost = Daily cost × 30

Pricing Reference (May 2026)

Model	Input	Output
Claude Sonnet	$3/1M	$15/1M
Claude Haiku	$0.80/1M	$4/1M
Claude Opus	$15/1M	$75/1M
GPT-5.5	$2/1M	$8/1M
GPT-5.5 Instant	$0.05/1M	$0.15/1M
Gemini 3.1	$2/1M	$12/1M
DeepSeek V4	$0.55/1M	$2.19/1M
DeepSeek V4 Flash	$0.14/1M	$0.28/1M

Cost Optimization Tips

For detailed cost optimization strategies, see Economics of AI. Key techniques include model routing, prompt caching, semantic caching, and batch processing.

When Costs Get High

Monthly Cost	Action
Under $10	Ignore, just use what you need
$10-100	Monitor, consider optimization
$100-1000	Implement routing/caching
Over $1000	Evaluate fine-tuning or self-hosting

ROI Calculation

Question: Is the cost justified?

Calculation:

Time saved per request × Value of your time × Requests/month
  vs
Cost per month

Example: Chatbot saves 5 minutes per conversation

5 min × $1/min =$ 5 value
100 conversations/month = $500 value
Cost: $2.10/month
ROI: 23,700% ✅