Skip to content

Free tools

AI Agent Cost & Markup Calculator

Work out what your agent runs cost, what to bill the client, and what you keep.

Free, no sign-up. Everything runs in your browser; nothing you type is sent to us.

Tokens per run

Prompt tokens served from the cache.

Tokens written to the cache (5-minute).

What you add on top of cost when billing the client.

Price to client per month

$83.20

Cost per run
$0.64
Cost per month
$64.00
Your margin per month
$19.20
Margin on price
23.1%

Model prices used

List prices in USD per million tokens, from each provider's pricing page. The eBRBe product uses the same table.

ModelInputOutputCache readCache writeSource
Claude Opus 5.5claude-opus-5-5$4.00$20.00$0.20$5.00Anthropicas of 2026-10-28
Claude Sonnet 5.5claude-sonnet-5-5$2.00$10.00$0.10$2.50Anthropicas of 2026-10-28
Claude Haiku 5.5claude-haiku-5-5Prompts over 100,000 tokens: $0.50 / $2.50$0.10$0.50$0.01$0.125Anthropicas of 2026-10-28
Claude Opus 5claude-opus-5$5.00$25.00$0.50$6.25Anthropicas of 2026-10-28
Claude Sonnet 5claude-sonnet-5$2.00$10.00$0.20$2.50Anthropicas of 2026-10-28
Claude Haiku 4.5claude-haiku-4-5$1.00$5.00$0.10$1.25Anthropicas of 2026-10-28

How it's calculated

Each token type is priced separately: input, output, cache reads and cache writes, using the model's list price per million tokens. Cost per run times runs per month gives your monthly cost; the markup gives the price to the client.

Prices are list prices for direct API use. Batch processing, regional endpoints and negotiated discounts change the real figure, so check your provider's invoice.

Billing agents to clients

Agencies bill AI work in a few ways: pass the cost through, add a markup for managing it, or absorb it in a retainer. Whichever you choose, put it in writing and show it on the invoice so the client can see what they pay for. Our guide, how to bill clients for AI agent costs, covers each model, markup versus margin, and what to put on the invoice.