Saarika V2.5
Saarika V2.5 transcribes speech in the spoken language. Its guide covers 11 languages and marks the release for upcoming deprecation, directing new transcription work to Saaras V4.
Deprecated API; check access details
Preserve the transcription objective
The guide documents automatic language detection and batch speaker diarization. The suggested replacement is Saaras V4 in transcribe mode, not its English translation mode.Saarika model and deprecation guide
Practical use
Practical guidance. A migration should preserve the expected output language. Check mixed-language calls, names and numbers explicitly; a readable English translation is not a substitute for the original-language record.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Status | Deprecated API; check access details |
| Access | API |
| Model identifiers | saarika:v2.5 |
| Lifecycle | Deprecation announced; no fixed withdrawal date on the cited pageSaarika model and deprecation guide |
Related models
A speech-recognition model for Indian languages and global English with normalized, verbatim, transliterated, code-mixed and translated output.
Launch date not confirmedAPI