Codesota · Models · LightOnOCR-2-1BLightOn8 results · 1 benchmarks
Model card

LightOnOCR-2-1B.

LightOnopen-sourceOCR / document VLMcheck model card3 current SOTA

Added from Papers With Code public API dump for olmOCR-Bench.

§ 02 · Benchmarks

Every benchmark LightOnOCR-2-1B has a recorded score for.

#BenchmarkArea · TaskMetricValueRankDateSource
01olmOCR-BenchVision & Documents · Document Parsingbase99.6%#3/42026-04-20source ↗
02olmOCR-BenchVision & Documents · Document Parsinglong-tiny-text91.4%#2/42026-04-20source ↗
03olmOCR-BenchVision & Documents · Document Parsingarxiv89.6%#1/52026-04-20source ↗
04olmOCR-BenchVision & Documents · Document Parsingtables89.0%#1/52026-04-20source ↗
05olmOCR-BenchVision & Documents · Document Parsingold-scans-math85.6%#1/42026-04-20source ↗
06olmOCR-BenchVision & Documents · Document Parsingmulti-column84.8%#2/42026-04-20source ↗
07olmOCR-BenchVision & Documents · Document Parsingold-scans42.2%#4/52026-04-20source ↗
08olmOCR-Bench—pass-rate83.2%#4/332026-01-20source ↗
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 03 · Strengths by area

Where LightOnOCR-2-1B actually performs.

Vision & Documents
1
benchmark
avg rank #2.0 · 3 SOTA
§ 06 · Sources & freshness

Where these numbers come from.

codesota-api
7
results
paperswithcode-public-api
1
result
1 of 8 rows marked verified. · first result 2026-01-20, latest 2026-04-20.