Genpio

Genpio Research

Original measurement published with its method and its raw data. The AI Livestream Selling Benchmark 2026 protocol is open before its results exist.

Original measurement, published with its method and its raw data. One study is open right now, and its protocol was published before any of its results existed.

Last updated: 2026-08-23

Buyers evaluating AI livestream platforms are currently choosing between vendor self-reports. Ours says a comment reaches a generated reply in under 200ms. A competitor's says it responds to every comment. Neither statement was measured by anyone who is not selling something, and neither is expressed in units the other can be compared in.

That is a gap in the category rather than only in our marketing, and it is what this programme exists to close.

How we intend to publish research

Protocol before results. A method settled after the numbers are in is a method that can be tuned, quietly, toward whoever paid for it. Publishing the protocol first is the only mechanism we know of that makes that visible, and it is a commitment that can be held against us if the results never appear.

Raw data, openly licensed. Summary tables are where errors hide. Every study publishes its measurements as CSV under CC BY 4.0, so anyone can recompute the summary and get a different answer if we got it wrong.

The harness, published. So the measurements can be reproduced against vendors we did not test.

Right of reply. Every vendor measured gets its own raw data before publication with a stated window to respond, and any correction or objection is published alongside the result whether or not we agree with it.

Where a competitor wins, that is the result. A benchmark whose sponsor wins every metric is a brochure. If ours wins everything, treat that as evidence the method is bad.

The model card for Genpio Voices names every training corpus with its id, licence, sample rate and hours, on a domain we do not control. The build and licensing record is on technology. Those are the same standard of disclosure this programme is trying to bring to the category's performance claims.

Pages in this section

  • The AI Livestream Selling Benchmark 2026 An open, pre-registered benchmark of AI livestream selling platforms, measuring latency, uptime, language coverage and cost per stream hour. This page publishes the methodology before the results exist, so the method cannot be tuned to flatter whoever ran it. Including us.