Legal
What we claim — and how to check it
Benchmarks, product pages and documentation on this site follow the rules below. If a statement ever seems to break them, tell us — that is a bug.
Last updated: 9 September 2026 · OPS2.AI LTD, United Kingdom, Company No. 17081809, registered address 71-75 Shelton Street, Covent Garden, London, WC2H 9JQ, United Kingdom. Contact: nlsys.io@proton.me.
Claim policy
Every claim is bound to something you can check
Published numbers are executed runs
Benchmark tables retain results with their stated protocol and version. Negative gain means increased error in that test. Identity/pass-through is an explicit status; a numerical zero alone does not identify the output or reason.
Models are judged in free-run
Compiled models are evaluated in closed-loop free-run: after the initial history, they receive their own previous predictions plus known inputs. This exposes error accumulation that one-step scores can hide.
Refusal is a feature
On near-clean input, a do-no-harm criterion backs the filter off to the raw signal. While a channel is still calibrating, the API flags it and may return a clean passthrough rather than an invented signal. Where the data does not support a claim, the model says so.
Support boundaries are stated
Every compiled model states its observed operating range and its policy for inputs outside that range. Its Passport also records limitations on the conclusions supported by the model.
Artifacts are verifiable
A compiled artifact carries a manifest, checksums and a Passport hash. Claims travel with the artifact, and the artifact can be checked — not taken on faith.
Status labels mean something
Product availability, price and current limits are shown in the customer account. Public pages distinguish released functionality from planned functionality; planned features are not described as current products.
What we do not claim
We do not claim improvement on every signal, causal discovery from a predictive graph, exact governing laws from arbitrary data, universal model validity, or safety-certified autonomous control. Benchmark conclusions apply only to the stated protocol and evidence.