Chat and support
Reusable instructions with short answers
Regular input
55%
Cache writes
5%
Cache reads
30%
Output
10%
Free tool
AI Token Cost Calculator
Choose an OpenAI or Claude API model, enter each token type in millions, and see the estimated cost instantly.
Regular input
$2.00
per 1M tokens
Cache write
$2.50
per 1M tokens
Cache read
$0.2000
per 1M tokens
Output price
$12.00
per 1M tokens
Estimated total
$14.00
GPT-5.6 Terra
Your token mix
Percent of total token volume, not percent of cost.
Regular input 50.0%
Cache writes 0.0%
Cache reads 0.0%
Output 50.0%
Cost breakdown
$2.00
Regular input
$0.0000
Cache writes
$0.0000
Cache reads
$12.00
Output
API prices verified: 2026-09-01.
USD estimate using standard first-party text-token rates, including 5-minute cache writes and cache reads. Anthropic 1-hour cache writes, batch, tools, images, audio, regional processing, and long-context rates are not included.
Typical token mix starting points
Illustrative planning patterns, not industry benchmarks. Measure production traffic and adjust for prompt length, response length, and cache reuse.
RAG and document Q&A
Retrieved context with strong cache reuse
Regular input
45%
Cache writes
5%
Cache reads
45%
Output
5%
Agents and coding
Tool loops with medium-length outputs
Regular input
50%
Cache writes
5%
Cache reads
30%
Output
15%
Content generation
Longer generated responses
Regular input
40%
Cache writes
5%
Cache reads
10%
Output
45%
Need a production AI workflow with predictable costs?
Urbano DX builds owned AI workflows with usage limits, caching, observability, and human review.