Neocodec 1.0.11
More capable transcription, reusable exports, speaker labels, and clearer cost and privacy controls.
What changed
- MAI-Transcribe 2 is now the default for new settings. Existing model choices remain in place. If a saved model is removed or unavailable, Neocodec asks you to choose a replacement before it uploads media.
- Transcripts can be saved as plain text, SRT, or WebVTT with optional speaker labels. Names and specialist words can be added as vocabulary hints, and the main window summarizes the saved choices.
- A completed transcript can be saved in another format without another upload. If saving fails, choose another folder and retry. Completed parts can also be saved as marked partial text after a later failure or cancellation.
- MAI-Transcribe 2 uses smaller lossless audio uploads in a format the model accepts. Neocodec checks word coverage before joining parts, keeps valid received text when optional timing is malformed, and preserves literal ampersands in SRT files.
- Each transcription row shows estimated and reported costs. Copy cost details includes usage notes and any receipt IDs returned with the result.
- Updated upload notices require renewed approval and explain that retention depends on the selected provider and your OpenRouter account settings. Neocodec does not guarantee zero retention.
Things to know
Speaker numbers restart in each audio part. Subtitle times include the source video's trim start. Provider timeouts can prevent completion, so review important transcripts against the recording.
Read the Privacy Policy, EULA, or contact support.