- RAID · ACL 2024
- Ed25519 · FIPS 186‑5
- Bitcoin anchored
Trust is a liability. Proof is an asset.
Every result is signed, and you can recompute it yourself. Need it on your own hardware? Deploy on-prem or air-gapped — then your text never leaves your network. Request a trial and run it on your own documents.
An accuracy claim you can defend
Every figure ships with the evidence to reproduce it. Your review board gets a number they verified — not a vendor’s marketing they have to accept.
Your text stays yours
Never stored, sold, or monetized. Deploy on-prem or air-gapped and your documents never leave your network.
Runs on your own CPUs
No GPU, no data center, no outside API in the detection path. Security by default, so you never have to think about it.
For your code, too
Engineering and security teams flag AI-written code in your environment; your source never leaves. See code detection →
We'll reach out shortly.
Thanks — your request is in. Someone from our team will contact you to set up your trial. In the meantime you can run a document through the tester.
Where institutions put Proofline to work
Academic-integrity review that survives an appeal
A signed, reproducible result a student — or a review board — can independently verify, instead of an opaque score.
Behind every flagged document is a person. Accuracy isn’t about being right — it’s about not being wrong.
Submission & content screening at volume
Batch-check manuscripts or contributed content through your own pipeline, with an audit trail you keep — no third-party service holding your text.
Air-gapped, procurement-ready
Runs entirely on your CPUs with no external calls in the detection path — the property a security review needs before anything is approved.
Non-English coverage the leaderboards skip
Full-document scoring in any language on the verifiable pipeline, where most public benchmarks report English only.
Know what’s AI-written in every pull request
Flag AI-generated code in review, CI, and audit — for security, accountability, and IP/licensing provenance. Runs in your environment; your code never leaves your network.
| What buyers ask | Typical AI detector | Proofline |
|---|---|---|
| Can we verify the accuracy claim? | Headline number, no method | Signed artifact you recompute |
| Does our text leave our network? | Sent to a hosted API | No by default — on-prem & air-gapped options |
| Is there generative AI in the loop? | Often another model | No |
| Reproducible on our own hardware? | No | Yes — Ed25519 + OpenTimestamps |
Comparison of properties, not a head-to-head accuracy claim. Every Proofline figure ships with its signed source — see the proof page.