Supported Languages
Valossa AI supports 23 languages across its various AI features, and there is also an experimental autodetect feature, which supports even more languages (set job language to autodetect). Language support varies by feature — not all features are available in all languages.
The default language for analysis is English (en-US). You can specify a different language in the media.language field of your new_job request.
Language Codes
| Language | Code |
|---|---|
| English (US) | en-US |
| German | de-DE |
| Spanish | es-ES |
| French | fr-FR |
| Italian | it-IT |
| Portuguese (European) | pt-PT |
| Portuguese (Brazilian) | pt-BR |
| Dutch | nl-NL |
| Swedish | sv-SE |
| Finnish | fi-FI |
| Ukrainian | uk-UA |
| Danish | da-DK |
| Norwegian (Bokmal) | nb-NO |
| Estonian | et-EE |
| Lithuanian | lt-LT |
| Latvian | lv-LV |
| Czech | cs-CZ |
| Polish | pl-PL |
| Greek | el-GR |
| Hungarian | hu-HU |
| Romanian | ro-RO |
| Russian | ru-RU |
| Kazakh | kk-KZ |
In-Video Metadata Features by Language
Languages are sorted by the number of supported features (most at the top).
| Language | Speech-to-text | Speech keywords & names | Video OCR | OCR content compliance | Scene IAB 2.2 (speech+OCR+visual) | Scene IAB 2.2 (visual only) | Bad language detection |
|---|---|---|---|---|---|---|---|
| English (en-US) | x | x | x | x | x | x | |
| German (de-DE) | x | x | x | x | x | ||
| Spanish (es-ES) | x | x | x | x | x | ||
| French (fr-FR) | x | x | x | x | x | ||
| Italian (it-IT) | x | x | x | x | x | ||
| Portuguese (pt-PT) | x | x | x | x | x | ||
| Portuguese BR (pt-BR) | x | x | x | x | x | ||
| Dutch (nl-NL) | x | x | x | x | x | ||
| Swedish (sv-SE) | x | x | x | x | x | ||
| Finnish (fi-FI) | x | x | x | x | x | ||
| Danish (da-DK) | x | x | x | x | x | ||
| Norwegian (nb-NO) | x | x | x | x | x | ||
| Hungarian (hu-HU) | x | x | x | ||||
| Romanian (ro-RO) | x | x | x | ||||
| Estonian (et-EE) | x | x | x | ||||
| Lithuanian (lt-LT) | x | x | x | ||||
| Latvian (lv-LV) | x | x | x | ||||
| Czech (cs-CZ) | x | x | x | ||||
| Polish (pl-PL) | x | x | x | ||||
| Ukrainian (uk-UA) | x | x | |||||
| Greek (el-GR) | x | x | |||||
| Russian (ru-RU) | x | x | |||||
| Kazakh (kk-KZ) | x | x |
English is the only language with full-modality Scene IAB 2.2 (combining speech, OCR, and visual analysis). All other languages use visual-only Scene IAB 2.2.
Video Categories and Summaries by Language
| Language | AV topics & categories | IAB 2.1 categories | IAB 2.2 overview (speech+OCR+visual) | IAB 2.2 overview (visual only) | LLM summarization, topics & genre |
|---|---|---|---|---|---|
| English (en-US) | x | x | x | x | |
| Finnish (fi-FI) | x | x | x | x | |
| German (de-DE) | x | x | x | x | |
| Spanish (es-ES) | x | x | x | x | |
| French (fr-FR) | x | x | x | x | |
| Italian (it-IT) | x | x | x | x | |
| Portuguese (pt-PT) | x | x | x | x | |
| Portuguese BR (pt-BR) | x | x | x | x | |
| Dutch (nl-NL) | x | x | x | x | |
| Swedish (sv-SE) | x | x | x | x | |
| Danish (da-DK) | x | x | x | x | |
| Norwegian (nb-NO) | x | x | x | x | |
| Ukrainian (uk-UA) | x | x | x | x | |
| Estonian (et-EE) | x | x | x | x | |
| Lithuanian (lt-LT) | x | x | x | x | |
| Latvian (lv-LV) | x | x | x | x | |
| Czech (cs-CZ) | x | x | x | x | |
| Polish (pl-PL) | x | x | x | x | |
| Greek (el-GR) | x | x | x | x | |
| Hungarian (hu-HU) | x | x | x | x | |
| Romanian (ro-RO) | x | x | x | x | |
| Russian (ru-RU) | x | x | x | x | |
| Kazakh (kk-KZ) | x | x | x | x |
English and Finnish are the only languages with full-modality IAB 2.2 overview categories (combining speech, OCR, and visual analysis). All other languages use visual-only IAB 2.2 overview.
Diarization Support
Speaker diarization (speaker identity separation) is in subscriptions that include speech recognition.
Notes
- Valossa adds supported features for languages according to commercial priorities. Contact Valossa sales if you need to know more about feature availability for a particular language.
- Language-specific feature availability is subject to change as new capabilities are added.
- Additional languages will be supported in the future.
- By default, speech is analyzed as English. Always specify the correct language code in your
new_jobrequest for non-English content.