Radar ·
Drama-3 added as streaming TTS in a speech-to-speech pipeline in a public repository
A pull request adds the Drama-3 model, from Fish Audio, as a streaming speech synthesis option in a speech-to-speech pipeline, with optional voice cloning and expressive intonation.
According to the pull request description, published on October 19, 2026, the change adds the option --tts drama3 to the speech-to-speech repository of a large AI model platform. The backend generates streaming speech, can be self-hosted, and offers optional voice cloning. Expressive intonations, such as whispering or laughing, are among the features mentioned. The source does not report a merge date, quality metrics or latency.
To verify the original material, consult the public repository indicated in the Radar record and check the pull request status, comments and tests before treating the feature as available. If your team uses AI to study this architecture, do not paste real customer audio, names or identifiers into external tools; use synthetic samples. This news does not mean that Rota Nacional offers speech synthesis with this model.