Loading benchmarks…
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 |
|---|---|---|---|---|---|---|
| English → GujaratiEN · English→GU · Gujarati · ગુજરાતી | 329 | Mean 5.85 25th pct. 5.44 · Median 5.82 · 75th pct. 6.19 | Mean 0.865 25th pct. 0.857 · Median 0.869 · 75th pct. 0.873 | Mean 8.9 25th pct. 8.3 · Median 8.8 · 75th pct. 9.8 | Mean 15.5 25th pct. 14.5 · Median 15.3 · 75th pct. 16.4 | Mean 0.998 25th pct. 0.983 · Median 0.990 · 75th pct. 1.012 |
| Portuguese → AmharicPT · Portuguese · Português→AM · Amharic · አማርኛ | 329 | Mean 5.07 25th pct. 4.88 · Median 5.56 · 75th pct. 5.80 | Mean 0.854 25th pct. 0.855 · Median 0.861 · 75th pct. 0.867 | Mean 3.8 25th pct. 3.7 · Median 3.8 · 75th pct. 3.9 | Mean 5.6 25th pct. 5.5 · Median 5.6 · 75th pct. 5.8 | Mean 0.804 25th pct. 0.786 · Median 0.795 · 75th pct. 0.806 |
| Russian → MalayalamRU · Russian · Русский→ML · Malayalam · മലയാളം | 329 | Mean 6.07 25th pct. 5.82 · Median 6.14 · 75th pct. 6.51 | Mean 0.766 25th pct. 0.753 · Median 0.774 · 75th pct. 0.778 | Mean 1.1 25th pct. 0.9 · Median 1.0 · 75th pct. 1.2 | Mean 4.3 25th pct. 4.0 · Median 4.4 · 75th pct. 4.8 | Mean 0.888 25th pct. 0.873 · Median 0.880 · 75th pct. 0.912 |
| Mandarin Chinese → AmharicZH · Mandarin Chinese · 普通话→AM · Amharic · አማርኛ | 329 | Mean 5.23 25th pct. 5.03 · Median 5.46 · 75th pct. 6.20 | Mean 0.829 25th pct. 0.822 · Median 0.834 · 75th pct. 0.840 | Mean 4.0 25th pct. 3.8 · Median 4.1 · 75th pct. 4.2 | Mean 5.6 25th pct. 5.3 · Median 5.6 · 75th pct. 5.8 | Mean 4.713 25th pct. 4.560 · Median 4.717 · 75th pct. 4.774 |
| Mandarin Chinese → BurmeseZH · Mandarin Chinese · 普通话→MY · Burmese · မြန်မာဘာသာ | 329 | Mean 5.39 25th pct. 5.06 · Median 5.32 · 75th pct. 6.00 | Mean 0.861 25th pct. 0.855 · Median 0.863 · 75th pct. 0.865 | Mean 1.8 25th pct. 1.5 · Median 2.1 · 75th pct. 2.3 | Mean 2.6 25th pct. 2.2 · Median 2.7 · 75th pct. 3.2 | Mean 3.056 25th pct. 2.865 · Median 3.069 · 75th pct. 3.269 |
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.