lightonai/LightOnOCR-3-1B
mean · 71.4 source-defined score; unit unspecified
Source-qualified evaluations, report statistics and safety observations.
mean · 71.4 source-defined score; unit unspecified
swe_bench_%_resolved · 57.6 source-defined score; unit unspecified
zh_t · 45.6 source-defined score; unit unspecified
non_latin · 67 source-defined score; unit unspecified
MathArena/aime_2026 · 92.7 source-defined score; unit unspecified
lhtb_solved · 3 source-defined score; unit unspecified
lexam_hard · 45.72 source-defined score; unit unspecified
mean · 83.17 source-defined score; unit unspecified
toolathlon_verified · 48.7 source-defined score; unit unspecified
general · 62.7 source-defined score; unit unspecified
category_macro_f1 · 0.488 source-defined score; unit unspecified
chart · 0.9 source-defined score; unit unspecified
diamond · 84.8484848485 source-defined score; unit unspecified
information · 29.28 source-defined score; unit unspecified
gpqa_diamond · 72.9 source-defined score; unit unspecified
text_content · 64.43 source-defined score; unit unspecified
de · 89.7 source-defined score; unit unspecified
mean · 46.17 source-defined score; unit unspecified
These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.