Finishing provider sign-in (or email link). This can take a few seconds on a slow connection.
Use the same Supabase account as speaktrue.cc (email, Google, Apple, or the marketing sign-in page).
Review before using SpeakTrue
SpeakTrue uses ElevenLabs as a third-party AI speech service. Before any app feature sends your personal data or user content to ElevenLabs, you must review this notice and give permission.
ElevenLabs is the third-party AI service used for speech processing.
You are responsible for the text, recordings, generated speech, cloned voices, and soundboard content you submit or create. Do not use SpeakTrue for unlawful, harmful, abusive, deceptive, impersonation, spam, robocalling, rights-infringing, high-risk, or provider-policy-violating content. SpeakTrue is provided as-is and service behavior may depend on provider availability and network conditions.
SpeakTrue collects account/session identifiers, submitted text, recordings, voice clone samples, generated/transcribed content, soundboard files and metadata, settings, usage data, and diagnostics only as needed for app functionality, personalization, reliability, and support.
The Privacy Policy identifies what data is collected, how it is collected, how it is used, retention choices, and third parties including ElevenLabs and Supabase. It also states that ElevenLabs provides its own privacy and security protections, and that SpeakTrue relies on those practices to provide the same or equal protection for user content processed by ElevenLabs.
Enter the email and password for your SpeakTrue account.
Enter up to 10,000 characters. Text over 5,000 requires Eleven Multilingual v2.
Your generated speech will appear here.
Quick generation supports up to 10,000 characters. Admin test accounts can switch to Long-form for lecture scripts, using the same voice and settings.
Record with your microphone or choose an audio or video file.
Review, edit, or send the result to Text to Speech.
Upload or record audio, choose the source and transcription settings, then select Transcribe. Supported formats include MP3, WAV, M4A, MP4, and other common audio or video formats. File size limit depends on provider and file processing limits. Enable speaker diarization when you need separate speakers. Enable audio event tagging when non-speech sounds matter. You can set defaults in Settings. Results remain editable and can be sent directly to Text to Speech.
Record your speech or choose an audio or video file.
Review the transcript and listen to the converted voice.
Upload or record speech, choose an output voice and model, then generate a clear spoken result. Share the generated audio by itself when you only need the converted clip. Reuse saved style presets and pronunciation glossary entries when supported by the selected model. The transcript and generated artifact remain available for TTS and Soundboard saving.
Open a category directly from this tab. Your clip playback, reorder, combine, and export flows stay in the same in-category panel.
saved_tts). Sign in with the same app session, then refresh.
Moves the clip at the first number into the slot of the second; clips at and below the target shift down. Use Save Order for drag-and-drop changes.
Estimated length: -
Strict mode: sign in to sync settings with your account.
Standard: record first, then transcribe and generate the translated voice after processing.
Turn-Based STS: recommended. Listen for one turn, then speak the completed transcript in the selected voice.
Custom terms apply to TTS and STT when a pronunciation entry exists. Entries are stored in your app database (same as iOS/Android glossary).
Create custom voice clones from your audio samples using AI technology
Record with a guided script or upload 1-10 clear audio samples. Each sample should be at least 30 seconds; 1-3 minutes total gives better results.
Choose a script and read it naturally. Pauses and small mistakes are okay.
Your browser will ask for microphone access. The recording stays in your browser until you attach it.
Drag and drop audio files here
or
Run sample-quality checks before creating your clone (same backend preflight used by iOS/Android).
Preflight not run yet.
Processing...
Select a category to save the clip to: