BlogLLM Benchmark Scores as Published, Side by Side
CoVoST: Results as Published
Every result the catalog holds for CoVoST: 3 models from 1 developer, read from 1 page. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.
Area Multimodal UnderstandingMaintainer Meta (Facebook AI Research)Introduced 2020Data version 2026103001, as of 2026-10-30
CoVoST 2 tests speech-to-text translation: the model hears speech and writes its translation. The corpus covers translation from 21 languages into English and from English into 15 other languages, built on Common Voice recordings. Publishers usually report BLEU, the standard translation score, but a model card that prints one number does not always name the metric.
CoVoST (metric not stated)
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Gemma 4 12BGoogle, released 2026-04-02 | 38.5 | Google, Gemma 4 model cardchecked 2026-10-15 | - |
| Gemma 4 E2BGoogle, released 2026-04-02 | 33.47 | Google, Gemma 4 model cardchecked 2026-10-15 | - |
| Gemma 4 E4BGoogle, released 2026-04-02 | 35.54 | Google, Gemma 4 model cardchecked 2026-10-15 | - |
Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.