Product

One tool for AI cost management, from "what does this cost" to "what did we spend".

Five pieces that share one pricing source: a directory, calculators, a usage workspace, budgets and a what-if comparison engine.

Directory

Every model's price, checked against the source

The starting point for AI cost management is knowing what things actually cost. The directory covers every major provider, links each figure to its source, and flags anything we couldn't independently verify instead of quietly presenting it as fact.

Browse the directory

Calculators

Turn a token volume into a dollar figure

Enter requests, input and output tokens, and get an exact cost from live pricing — for one request or a projected month. No spreadsheet, no reading a provider's tier structure to find the rate that applies to you.

Open the calculator

Example, not your data

Model

Amazon Nova 2.0 Pro

Tokens

50K in / 12K out

Estimated cost$0.2008

Workspace · projects & CSV import

Bring in usage the way you already have it

Create a project per product, team or client, log usage by hand, or import a CSV export straight from a provider's usage dashboard. Every imported row is re-priced from the same verified rates in the directory, so cost stays consistent everywhere it's shown.

See what's included in Pro
usage-export-june.csv3 rows parsed
  • 2026-06-24 GPT-4o 84.2K in / 19.1K out$0.4015
  • 2026-06-25 Claude Sonnet 5 51K in / 14.8K out$0.25
  • 2026-06-25 Gemini 3.5 Flash 210K in / 38.5K out$0.6615

Example file. Cost is computed from live pricing at import time, matching the directory.

Workspace · budgets & alerts

Know before you're over, not after

Set a monthly budget per project and a threshold to alert at. Meterlark evaluates every new entry against it, so a runaway job or a pricing change shows up as a warning instead of a surprise on next month's total.

See what's included in Pro

Budgets & alerts

Example workspace · Pro feature

  • Support bot$602.71 / $450.00

    Overalert sent at 134% of budget

  • Content pipeline$502.26 / $520.00

    Nearingalert sent at 97% of budget

  • Internal search$1,130 / $700.00

    Overalert sent at 161% of budget

Workspace · what-if

Price your real usage against every other model

Take your actual month-to-date usage and run it through every model in the directory. This is where AI cost management stops being theoretical: it's the exact answer to "what would switching actually save us".

Compare two models

What if we switched?

Example workspace · same month-to-date usage, every model

  • Claude Haiku 4.5Cheapest$2.00
  • Amazon Nova 2.0 Pro$3.66
  • Claude Sonnet 5Current$4.00
  • Claude Opus 4.8$10.00
  • Claude Fable 5$20.00

Start with the free half.

The directory and calculators need no account. Sign up when you want the workspace tracking real usage.

Start free