Available now

Chatterbox HD: text to speech

Chatterbox HD reads your text in one of its preset voices or in a voice cloned from a short sample, with control over emotion.

Model ID

Input

Fields marked * are required. Review the price before generating.

Text to synthesize into speech.

Example. Edit it or write your own.

URL to the audio sample to use as a voice prompt for zero-shot TTS voice cloning. Providing a audio sample will override the voice setting.

Uploading a file needs an account. Create account

MP3, WAV · up to 50 MB

Additional settings

The voice to use for the TTS request. If neither voice nor audio are provided, a random voice will be used.

Controls emotion exaggeration. Range typically 0.25 to 2.0.

Controls the randomness of generation. Range typically 0.05 to 5.

Price$0.07 / run

You sign up first. Your settings are kept.

About this model

Overview

Chatterbox HD turns text into speech. Choose a preset voice such as Aurora, Carl or Siobhan, or upload an audio sample and it speaks in that voice instead. With neither, it picks a random voice.

Exaggeration sets how emotional the read sounds and guidance how closely it follows the voice. The high quality option returns 48kHz audio instead of 24kHz and takes longer.

What you can build

  • Voice-overs in a preset voice for videos
  • Narration in a voice cloned from a short sample
  • Dramatic character lines for games and stories
  • The same line read by several preset voices

Prompting tips

  • For a dramatic read, lower guidance to about 0.3 and raise exaggeration to 0.7 or more.
  • If the sample speaker talks fast, a lower guidance value helps the pacing.
  • Keep the seed fixed while you try settings so only one thing changes at a time.

What to check

Listen for names, pronunciation, pauses, clipping and the intended tone before publishing.

Inputs

  • Text (required): Text to synthesize into speech.
  • Voice (optional): The voice to use for the TTS request. If neither voice nor audio are provided, a random voice will be used. Options: Aurora, Blade, Britney, Carl, Cliff, Richard, Rico, Siobhan, Vicky.
  • Audio (optional): URL to the audio sample to use as a voice prompt for zero-shot TTS voice cloning. Providing a audio sample will override the voice setting.
  • Exaggeration (optional): Controls emotion exaggeration. Range typically 0.25 to 2.0.
  • Cfg (optional): Classifier-free guidance scale (CFG) controls the conditioning factor. Range typically 0.2 to 1.0. For expressive or dramatic speech, try lower cfg values (e.g. ~0.3) and increase exaggeration to around 0.7 or higher. If the reference speaker has a fast speaking style, lowering cfg to around 0.3 can improve pacing.
  • High Quality Audio (optional): If True, the generated audio will be upscaled to 48kHz. The generation of the audio will take longer, but the quality will be higher. If False, the generated audio will be 24kHz.
  • Seed (optional): Useful to control the reproducibility of the generated audio. Assuming all other properties didn't change, a fixed seed should always generate the exact same audio file. Set to 0 for random seed.
  • Temperature (optional): Controls the randomness of generation. Range typically 0.05 to 5.

Output and limits

  • Sung or spoken words can come out unclear or differ from your text.
  • Exact length, tempo and key are not guaranteed; listen to the whole result before you use it.
  • Prompts that name real artists or protected songs can be refused.
  • Some prompts or outputs can be refused by safety checks.

Price guide

  • up to · 1,000 ASCII characters: $0.07 / run