interactive demo

Price an agent workload

Pick a real public dataset, or upload your own trace export, and see a quarter of weekly AI bills bounce around, alongside the flat number we'd write to guarantee certainty. Everything runs in your browser, so your data never leaves this page.

2,145 units

A CSV with a header row and one row per unit, or one row per call. Costs must already be in US dollars, and tokens are not read.

Cost, required
a column named cost_usd, cost, total_cost, total_cost_usd, usd or cost_per_unit
Unit id, optional
a column named unit_id, task_id, instance_id, trace_id, session_id or ticket_id. With one, rows are added up per id, so a per-call export prices correctly. Without one, every row counts as a unit.
Everything else
ignored. Rows with a missing or zero cost are skipped, and 10 usable units is the floor.
task_id,cost_usd,timestamp
tkt_1041,0.38,2026-08-02T14:11:09Z
tkt_1042,1.94,2026-08-02T14:12:31Z
Download a 60-unit sample CSV
mean $0.84one fixed price $0.94$0.00$0.50$1.00$1.50$2.00
6,306 units, priced by the engine on this page. Each bar is how many units cost that much, the axis is dollars per unit cut just past the 99.5th percentile, and the most expensive unit in the sample cost $2.84. Units past the line cost more than the fixed price, and on a Second Moment contract they land on our book instead of yours.

A quarter of weekly bills (same workload, two ways):weeks simulated from the dataset's own tasks with week-to-week mix shift

what you pay · flatyour average week
Red weeks cost more than the flat price: 1 of 13 here. Grey weeks cost less, and that gap is our margin. Now try forecasting which weeks would be red.
your average unit
$0.84
your worst units
$2.10
the worst 1% cost 2.5× the average
with second moment
$0.94 per unit, fixed
certainty costs $0.09/unit, 11% above your average.
your average week
$1,806
at 2,145 units/week
your worst simulated week
$2,143
19% over average
with second moment
$2,006 /week, every week
certainty costs $199/week

Other contract shapes:

If your workload or appetite calls for something else, the same distribution prices these contract shapes:

Three prices instead of one

One flat price makes easy units subsidize hard ones. If your task scope varies widely, we can quote Small/Medium/Large instead, each unit assigned a tier at intake, each tier priced as its own book.

TierShareMeanPrice
Small50%$0.52$0.56
Medium40%$1.02$1.10
Large10%$1.76$1.89

Tiers cut at the p50/p90 of your own distribution. Live tiers come from task features at intake.

Or start with just the cap

Not ready for a fixed price? Second Moment passes through costs up to an agreed threshold (plus a small fee) and absorbs anything above. This is where most contracts will start.

ThresholdFee / unit
$1.46 (p90)$0.0372
$1.69 (p95)$0.0156
$2.10 (p99)$0.0024

Example thresholds set at the p90/p95/p99 of your own units. Fee = 1.25 × the expected amount we absorb above the threshold.

Later: fixed price per outcome

Pay only for resolved units, for agents that have a well-defined terminal state. It needs your resolution rate measured on a carried book first, and the price carries the failure and correlation load (roughly 1.4–1.8× the per-unit rate), so it fits workloads whose value per outcome clears it. Cap → fixed per unit → fixed per outcome. Each step naturally funds the data that prices the next.

Show the math (how the number is built):

Why is the price what it is?

the quote
$0.94 per unit, fixed

mean $0.84 + risk load $0.02 (weekly aggregate q99 at 2,145 units) + 8.1% load = 4% base + 4.1% variance-linked (σ/μ 0.51) · Coding agent · 6,306 issues (SWE-rebench, public)

What do your units cost today?

Your cost distribution n=6,306 · Z=0.97

Unit costUSD× the average
the average unit$0.841.0×
p95$1.692.0×
p99$2.102.5×
the worst unit in the sample$2.843.4×
the fixed price$0.941.1×
mean $0.84, p95 $1.69 (2.0×), p99 $2.10 (2.5×). Tail beyond the sample blended with the code-tasks prior at weight 0.03.

Which weeks would we have lost?

Backtest: 13 weeks, blue we earn, red we absorb

$0$115$229worst $168
Base replay: you pay $26,074 flat vs $23,449 variable · Second Moment margin $2,624 · 0 losing week(s). At volume, week-to-week noise diversifies. Try the stress tabs.

The full math: formula and calibration, for your team →