Automated translation metrics summarized for 334 language directions.
Sources — Source segments come from curated, public-facing parallel corpora and benchmark-style collections mixed for broad coverage. They are not model-generated: originals and references are assembled upstream as parallel text only; this dashboard does not redistribute raw corpora.
Mixed parallel corpora — Parallel segments spanning diverse genres and corpus scales. Each item includes a reference translation that is certified and human-verified—not produced by the models being benchmarked.
Language pairi | Rankingi | Fluencyi | COMETi | BLEUi | sacreBLEUi | Len ratioi |
|---|---|---|---|---|---|---|
| Modern Standard Arabic → AmharicAR · Modern Standard Arabic · العربية الفصحى→AM · Amharic · አማርኛ | 171 | Mean 4.65 25th pct. 4.19 · Median 4.99 · 75th pct. 5.54 | Mean 0.827 25th pct. 0.828 · Median 0.834 · 75th pct. 0.841 | Mean 3.8 25th pct. 3.5 · Median 3.7 · 75th pct. 4.1 | Mean 5.6 25th pct. 5.0 · Median 5.6 · 75th pct. 5.9 | Mean 0.921 25th pct. 0.901 · Median 0.917 · 75th pct. 0.936 |
| Modern Standard Arabic → Egyptian ArabicAR · Modern Standard Arabic · العربية الفصحى→ARZ · Egyptian Arabic · مصري | 79 | Mean 7.63 25th pct. 6.89 · Median 7.65 · 75th pct. 8.38 | Mean 0.777 25th pct. 0.763 · Median 0.782 · 75th pct. 0.799 | Mean 20.0 25th pct. 13.5 · Median 21.7 · 75th pct. 26.3 | Mean 30.5 25th pct. 30.1 · Median 32.0 · 75th pct. 33.4 | Mean 0.992 25th pct. 0.969 · Median 0.998 · 75th pct. 1.012 |
| Modern Standard Arabic → AzerbaijaniAR · Modern Standard Arabic · العربية الفصحى→AZ · Azerbaijani · Azərbaycan | 248 | Mean 5.51 25th pct. 5.01 · Median 5.57 · 75th pct. 5.91 | Mean 0.809 25th pct. 0.789 · Median 0.818 · 75th pct. 0.827 | Mean 7.6 25th pct. 6.4 · Median 6.8 · 75th pct. 8.7 | Mean 14.2 25th pct. 12.6 · Median 13.5 · 75th pct. 16.3 | Mean 0.893 25th pct. 0.863 · Median 0.887 · 75th pct. 0.922 |
| Modern Standard Arabic → BengaliAR · Modern Standard Arabic · العربية الفصحى→BN · Bengali · বাংলা | 171 | Mean 6.59 25th pct. 6.25 · Median 6.83 · 75th pct. 7.01 | Mean 0.818 25th pct. 0.805 · Median 0.822 · 75th pct. 0.827 | Mean 10.3 25th pct. 6.7 · Median 10.8 · 75th pct. 13.3 | Mean 12.4 25th pct. 9.5 · Median 12.9 · 75th pct. 15.1 | Mean 1.161 25th pct. 1.135 · Median 1.162 · 75th pct. 1.187 |
| Modern Standard Arabic → GermanAR · Modern Standard Arabic · العربية الفصحى→DE · German · Deutsch | 125 | Mean 7.05 25th pct. 6.75 · Median 7.08 · 75th pct. 7.52 | Mean 0.780 25th pct. 0.769 · Median 0.783 · 75th pct. 0.792 | Mean 10.8 25th pct. 8.5 · Median 10.7 · 75th pct. 12.2 | Mean 14.8 25th pct. 12.3 · Median 14.8 · 75th pct. 16.1 | Mean 1.449 25th pct. 1.426 · Median 1.438 · 75th pct. 1.467 |
This release reports translation quality by language pair: medians and spread of automated scores (fluency, COMET, BLEU, sacreBLEU, length ratio) aggregated across evaluated directions. Run metadata describes recipes, metrics, and segment counts.