Unlimited-OCR-2: Baidu open-sources an OCR model for unlimited-horizon parsing
Comments on “Unlimited-OCR-2: Baidu open-sources an OCR model for unlimited-horizon parsing”
3 judges. Scored by all three judges. Only the day’s top stories get the claim check. How far stories are checked
Baidu has open-sourced Unlimited-OCR-2, a document parsing model whose new Sparse Reference Sliding Window Attention (SR-SWA) mechanism compresses attention over very long sequences to a fixed working set of about 512 tokens, so per-step decoding latency and memory stay roughly constant as documents grow.
Why it ranked
No. 92 of 155 with a score of 0.59. The panel rated it highest on new information (0.95) and technical depth (0.85), lowest on audience breadth (0.49).
- AI judgeJudge 2claude-haiku-5-5
- Open-source OCR model release with technical detail on attention and serving setup, a benchmark, and a paper. Substantive for AI builders, though niche in audience; claims of SOTA rest on the authors' own benchmarks.
- AI judgeJudge 3gpt-5.6-luna
- A technically detailed open model release with reproducible inference and serving instructions, addressing ultra-long document OCR. Its benchmark and efficiency claims are notable, though “unlimited” framing likely overstates practical bounds.
Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.
Score
0.59total under rubric v2.2
- Significance0.59
- Reader interest0.64
- Audience breadth0.49
- Practical relevance0.73
- Substantiated0.76
- Technical depth0.85
- New information0.95
- Counts against: Promotional0.12
- Counts against: Overclaim0.41
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| Significance | 0.65 | 0.40 | 0.72 | 0.59 | 0.32 | +0.24 | +0.141 |
| Reader interest | 0.54 | 0.50 | 0.88 | 0.64 | 0.38 | +0.26 | +0.166 |
| Audience breadth | 0.44 | 0.40 | 0.62 | 0.49 | 0.22 | +0.13 | +0.063 |
| Practical relevance | 0.81 | 0.60 | 0.78 | 0.73 | 0.21 | +0.22 | +0.160 |
| Substantiated | 0.89 | 0.60 | 0.79 | 0.76 | 0.29 | +0.05 | +0.038 |
| Technical depth | 0.99 | 0.70 | 0.86 | 0.85 | 0.29 | +0.05 | +0.043 |
| New information | 0.95 | 0.90 | 0.99 | 0.95 | 0.09 | +0.05 | +0.047 |
| Promotional | 0.10 | 0.15 | 0.12 | 0.12 | 0.05 | −0.19 | −0.023 |
| Overclaim | 0.46 | 0.40 | 0.38 | 0.41 | 0.08 | −0.11 | −0.045 |
| Sum of contributions | +0.590 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.590 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.
<a href="https://rubric.hrih.si/editions/2026-10-28/92/"><img src="https://rubric.hrih.si/badge/2026-10-28/92.svg" alt="Ranked #92 on Rubric, 28 Oct 2026" width="245" height="28"></a>Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 10 pts, not on front page
Recorded for context only, checked 28 Oct 2026, 06:47 UTC. It played no part in the ranking.