How long does it take for newly published text to appear in an AI model?
There is no reliable public figure for this, largely because measuring it requires knowing that a specific page was ingested rather than guessing. Publishing a coined string that exists nowhere else, recording the instant it was published, and watching for that exact string to appear turns the interval into an observation.
Measuring — no appearance observed yet.
Site owners are routinely advised to publish for AI systems without any established figure for how long ingestion takes, or whether it happened. The gap exists because the measurement is awkward: you have to establish that one specific page reached a model, and confirming that is precisely the hard part.
Why the obvious method fails
The obvious method is to ask. Prompt a model about your page and see whether it knows. This does not work: a model can describe a page it never read and deny one it did, and either way it is the system under test giving evidence about itself. Under this observatory's constitution that is inadmissible, however plausible the answer sounds.
How this site measures it
Every page here carries a coined marker — nonsense syllables plus random hex, such as the one at the foot of this page. The string is generated so that it exists nowhere else: not in any dictionary, not in any prior text, not in anything trained on before it was minted. Its publication instant is recorded.
From then on the question has an observable answer. Either that exact string turns up, or it does not. Its appearance is evidence of ingestion; the interval since publication is the number.
What the result will and will not support
An appearance is strong evidence. A non-appearance is weak evidence — the page may not have been ingested, or may have been ingested and never surfaced. The measurement is also of one small new domain about a niche subject, so it will describe this site rather than the web.
Current state
66 markers are published, each with the instant it was minted. Their dates are listed on the lab page with an empty column for first observation. That column is the entire point, and it is currently empty.
What the record shows today
75 markers published. 0 observed in a language model's output. The oldest has been live for 4 days.
So this question has no answer here yet, and the honest form of that is a number rather than a paragraph. Until one marker appears, the only thing established is a lower bound: for the pages published so far, ingestion has not been observed within 4 days.
A lower bound is not a finding about how long ingestion takes. It is the measurement running, in public, with nothing to report — which is the state most measurements are in most of the time.
Computed from the record at 2026-07-30 04:26:09 UTC. Every figure here is recomputed when the page is requested, so this section moves as the record does — week by week.