Has SoMark confirmed it?

No. SoMark has not confirmed it as of Oct 28.

Did SoMark open-source a 1.3B OXR document parser claiming top scores on three OCR benches?

ReportedLast checked Oct 28, 2026

SoMarkFirst seen Oct 27

Reported by QbitAI; removed from DailyCited /d/2026-10-28/ per product stamp 030948 (daily keeps Official only).

Evidence timeline · 1 source · newest first

Oct 28
Unconfirmed → Reported
Oct 27
Reported
QbitAI reports that startup SoMark open-sourced OXR (Optical Everything Recognition), a 1.3B document parser that the team says leads OmniDocBench, olmOCR-Bench, and DocParsingBench and parses faster than PaddleOCR-VL-1.6 and MinerU2.5-Pro in its tests. OXR targets 13 element types with Markdown/JSON output and a recursive self-improvement post-training loop.
Source: QbitAI

Quick answers

How we label rumors

Unconfirmed: Circulating with at least one public source we can link to; no official confirmation.
Reported: Reported by a news outlet; not confirmed by the company.
Confirmed: Confirmed by the company or another official source.
Debunked: Denied by the company, retracted by the original source, or shown to be false by the evidence.

We record what sources show, with dates. We don't predict or score how likely a rumor is.