Turn a recording into a new voice, one private WAV at a time.
CharaVox transcribes your recording, then speaks the transcript with a preset, community, or custom voice. It is a file-based re-dubbing workflow, not a real-time Discord microphone.
Upload a clean single-speaker recording or capture one in your browser. Fun-ASR creates the transcript, then the selected CharaVox voice synthesizes a new WAV. You can review the transcript and create a separately billed redo after editing it.
Because this workflow rebuilds speech from text, the output can have different timing, pace, and emotion. Background music, room sound, exact duration, lip sync, singing, and live voice routing are not preserved.
Drop your audio file here
MP3, WAV, M4A, WebM, OGG and FLAC
Maximum audio duration depends on your plan
Capabilities
Upload or record
Use WAV, MP3, M4A, WebM, OGG, or FLAC speech, or record directly from your browser microphone.
Automatic transcription
Fun-ASR converts clean single-speaker audio into editable text before the new voice is generated.
Three voice libraries
Choose a compatible platform preset, community voice, or one of your own authorized custom voices.
Private WAV result
Compare source and result, download the private WAV, and delete the job and its stored audio when finished.
Why creators choose it
Review the recognized words before editing and redoing a result.
Use the same managed, community, and custom voices available in CharaVox Studio.
Keep source and generated audio private with plan-based retention and owner-only playback.
Pay a predictable 4 credits for each verified second of source audio.
Use cases
Frequently asked questions
No. CharaVox processes an uploaded or recorded file through speech recognition and text-to-speech. It does not install a virtual microphone or change live calls, streams, Discord chat, or game audio.
No. The result is natural re-dubbing from the recognized text, so timing and delivery can change. Background music, room tone, singing, exact lip sync, and the original emotion are not preserved.
You can choose a compatible system preset, supported community voice, or your own authorized custom voice. CharaVox validates the voice and its available TTS model again when each job starts.
A job reserves and settles 4 credits per verified second of source audio. Free, Creator, and Pro jobs may contain up to 30 seconds, 5 minutes, and 10 minutes respectively.
After a result is ready, you can edit its non-empty transcript and create a new redo job. The redo skips speech recognition but is billed again using the original source duration.
Related AI tools
Speech to Text
Upload audio or record in your browser and turn speech into editable text with timestamps, speaker labels, and TXT or SRT exports.
ExploreAI Voice Generator
Turn any script into natural AI speech in 50+ languages and 27,000+ community voices. Pick a voice, paste your text, and download studio-grade voiceover in seconds.
ExploreVoice Cloning
Upload a 30-second audio sample and create a custom AI voice model. Generate speech in 6 languages with license-safe watermarking and commercial-use rights.
ExploreTikTok Voice Changer
TikTok voice changer with pitch-up, deep, robot, chipmunk and 200+ voice effects. Upload your audio or use TTS — get a transformed TikTok-ready voice in seconds.
ExploreCreate a new voice from recorded speech
Sign in to upload or record, choose an authorized voice, and generate a private WAV through the Fun-ASR and TTS workflow.