Powered by x402

You don't know what your AI subscription actually costs you.

We do — down to the cent, before it happens and after. Pay per request with a hard cap, or unlock a time pass when a burst of work makes more sense.

Estimated cost$0.04 – $0.06
Maximum charge
$0.08
Billing
Pay per request
Est. execution
~2.4s

You'll never pay more than $0.08 for this run.

Receipt

Jul 15, 2026 · 1:06 PM

txn_8f3a1c

Scout

Competitive landscape for micro-billing infra

  • Search · web queries$0.016
  • OpenAI · GPT-5 tokens$0.024
  • Embeddings · vector ops$0.012
Total charged$0.052
Execution 2.4sEstimated $0.04 – $0.06Settled via x402

Humans pay by time.
Machines pay by usage.

How it works

Three steps, every time — no matter which agent or which billing model.

01

Pick an agent

Browse by category, see ratings and runs, and check which billing models the creator has enabled — usage, passes, or both.

Sc

Scout

Deep web research

Featured
02

Approve the estimate

See the estimated cost and the hard maximum before anything runs. Nothing charges without your explicit approval.

Maximum charge$0.08
03

Get a receipt

Every run closes with an itemized breakdown — per-provider cost, execution time, and exactly where your money went.

Total charged$0.052

One price for everyone was never going to work

A flat monthly subscription overcharges casual use and undercharges heavy use. Adaptive billing charges what each task actually costs.

Traditional AI

Monthly subscription only

Jatah Ai

Usage, 24h, 7d, or 30d

Transparent per-task cost
Hard maximum charge
Approval before every charge
Pay only for what you use
Short-term access (24h / 7d)
Itemized receipts

Who saves, and how much

Weekend builder

One hackathon, ~40 runs

Subscription

$20.00

Jatah Ai

$2.10

Daily power user

~300 runs a month

Subscription

$20.00

Jatah Ai

$16.80

Heavy sprint week

Deadline week, constant use

Subscription

$20.00

7 Day Sprint

$9.00

A tier, a token allowance, no per-call tracking

Basic, Standard, or Premium — pick a model tier and a token allowance, available for a 24 Hour window or a full Week. Still no auto-renewal, still no surprise charges.

Basic

Everyday utility calls — fast, cheap, always on.

24 Hour

$1.50

1 Week

$5.00

8M tokens included

Per window — resets when a new pass starts.

  • Claude Haiku 4.5
  • GPT-5 Mini

Most teams start here

Standard

Most teams' daily driver — reasoning without the frontier price.

24 Hour

$3.50

1 Week

$10.00

20M tokens included

Per window — resets when a new pass starts.

  • Claude Sonnet 5
  • Gemini 2.5 Pro

Premium

Frontier models for the runs that can't afford to be wrong.

24 Hour

$6.00

1 Week

$16.00

40M tokens included

Per window — resets when a new pass starts.

  • GPT-5
  • Claude Opus 4.8

Go over your token allowance and the run just falls back to usage billing at the model's normal rate — never blocked, never a surprise.

Or pay for a window, not a task

Time passes work like a day pass at a gym — access for a window, set per agent by its creator, priced from $1.50.

24 Hour Pass

from $1.50

One deadline, one hackathon, one big push. Unlock a day of unlimited runs on the agent you need.

7 Day Sprint

from $7.00

A sprint week, a launch, a project deadline. Run as much as the week needs, then it's done.

30 Day Pro

from $18.00

For agents you lean on constantly — a month of access at a fraction of what per-request usage would cost.

Building an agent? You decide how it's billed.

Enable pay-per-request, time passes, or both. Set your own estimate ranges and caps. Payouts settle via x402 — no invoicing, no chasing.

Open Creator Studio

Command palette

Jump to a page, find an agent or model, or run an action