Build a conversation.
Paste or upload a JSON conversation to generate your audio. Use the Manual conversation tab when you want to edit individual turns.
Shared audio library
Every completed conversation is shared with all signed-in users. Press Play to listen here, or download WAV or MP3. Only the creator can delete a recording.
Import a JSON conversation
Upload a .json file or paste a version 1.0 conversation. Import replaces the current bubbles; Undo import restores the previous draft.
Up to 20 turns, 5,000 spoken characters (including tags), and a 256 KB JSON file.
Conversation & API info Eleven v4
Live settings for the next generation. Each bubble is generated separately, then joined into one recording.
- Settings
- Loading server configuration…
What can be changed across the conversation?
- Connected now
- Turn order, voice, language, script, stability and similarity; download either WAV or MP3.
- App configuration
- Model, character limits, sample rate, channels and MP3 bitrate. Per-turn silence is editable. V4 is fixed as the generation model.
- Additional API options · not connected
- Output codecs: MP3, WAV, PCM, Opus, μ-law and A-law. Some quality levels require higher plans. Logging defaults on; disabling it requires Enterprise. Legacy latency optimization is deprecated.
- Alternative generation modes
- Native multi-speaker dialogue, streaming and timestamped output require separate endpoint integration.
Your ElevenLabs website settings are separate from Aurnovem’s request settings. This app sends its own per-bubble values. V4 supports stability and similarity; pace and emotion can be directed in the script. The old 0.8× speed switch is no longer applied.
Edit the loaded conversation here. Re-importing JSON replaces these manual edits.
Completed audio saves automatically to the shared library.
Checking setup…