Scribeline

Proof: three real runs, on the record

One real quote from photos to a sent price, one run it refused to price, and what a run costs. Every figure here is measured from a logged run, not marketing.

A real quote, start to finish

Q-2026-0061

Recorded — a real run from 2026-07-17.

  1. 1.A customer sent photos and a few words.
  2. 2.The machine read the photos into 6 scope lines and asked 6 questions instead of guessing.
  3. 3.Every one of the 6 lines was priced from the owner's own book — $1,872.00 in all.
  4. 4.The owner reviewed it and approved it.
  5. 5.The branded quote went to the customer.
  6. 6.The customer accepted it.

What it asked instead of guessing:

Photos to a sent quote: 24.33 minutes.

The run where it said no

Q-2026-0076

What it read from the photos and words:

Repaint the exterior siding on this wall — the paint is faded and peeling in places.

The questions that blocked a price:

Why it stopped, in its own words:

No total. Nothing was sent.

A month of quoting, in one sitting

40 real jobs ran through the machine in 49 minutes of continuous run time — the volume a small shop quotes in a month, watched end to end.

The machine turned the photos into work lists and handed 92% of the lines back to the owner as measurement questions. It read at scale and still refused to guess.

The owner then supplied the measurements and drove 5 of them to sent quotes — $4,009 in all. The rest waited in the inbox, like a real Monday.

What a run costs

A run costs 25¢, drawn against an owner-funded test budget of $10.

The runs on this page were read by a paid cloud model. A private build reads on your own machine instead, where the running cost is about the electricity it draws.

A visitor never pays inside the app. There is no checkout here.

Back to the front doorDrive it yourself