Voice · Resemble AI
Chatterbox HD
Resemble AI's expressive voice model, here in private mode. Pick it for expressive narration that stays private.
Private High costWhat sets it apart
Chatterbox is Resemble AI's MIT-licensed voice model with an adjustable emotion exaggeration setting and built-in watermarking. Here it runs in private mode, so expressive output stays off third-party logs.
- Emotion exaggeration control
- MIT license, imperceptible watermark
- Voice cloning from 5 to 20 s of audio
Recommended for: read in french without leaving private mode.
Strengths and limits
Where it shines
- Private mode: nothing is kept
- Designed for emotional range, from flat to dramatic
Think twice when
- 9 voices only
- Mid-range price per character
- 9 voices, WAV output
- Up to 5,000 characters per request in Sub Rosa
Use cases
Dramatic audiobook scene
Read this passage with rising tension, as if the narrator is afraid.
The expressiveness setting suits emotional range.
Private character voices
Voice the villain of my short story, slow and menacing.
Private mode keeps drafts off third-party servers.
Against its rivals
- ElevenLabsElevenLabs has far more languages and polish, but does not offer private mode.#1 of 10 on Artificial Analysis Speech Arena (1,330 Elo)
- Orpheus TTSOrpheus is also private and expressive, but English only.
- KokoroKokoro is much cheaper and also private, but less dramatic.#10 of 10 on Artificial Analysis Speech Arena (1,064 Elo)
Version history
Each version, dated, with what it changed over the one before.
- Chatterbox HDLatest
- Introduced Chatterbox: open-source speech with zero-shot voice cloning and an emotion exaggeration control
Benchmarks
How to read these scores: what a benchmark tells you.
Artificial Analysis Speech Arena: #7 of 10 models measured in the catalog. People listen to the same sentence read by two voices, without knowing the models, and pick the more natural one. An Elo rating: higher sounds more natural to more people. 100 points apart means the higher-rated model wins about 64% of duels.
- 1. ElevenLabs TTS v4 TurboElevenLabs1,330
- 2. ElevenLabs TTS v4ElevenLabs1,329
- 3. Gemini 3.1 Flash TTSGemini 3.1 Flash TTS1,209
- 4. ElevenLabs TTS v3ElevenLabs1,178
- 5. Gradium TTSGradium TTS1,151
- 6. MiniMax Speech-02 HDMiniMax Speech-02 HD1,114
- 7. Chatterbox HD (Resemble AI)Chatterbox HD1,102
- 8. ElevenLabs Turbo v2.5ElevenLabs1,097
- 9. ElevenLabs Multilingual v2ElevenLabs1,095
- 10. Kokoro Text to SpeechKokoro1,064
See the numbers
| Model | Elo |
|---|---|
| 1. ElevenLabs TTS v4 Turbo | 1,330 |
| 2. ElevenLabs TTS v4 | 1,329 |
| 3. Gemini 3.1 Flash TTS | 1,209 |
| 4. ElevenLabs TTS v3 | 1,178 |
| 5. Gradium TTS | 1,151 |
| 6. MiniMax Speech-02 HD | 1,114 |
| 7. Chatterbox HD (Resemble AI) | 1,102 |
| 8. ElevenLabs Turbo v2.5 | 1,097 |
| 9. ElevenLabs Multilingual v2 | 1,095 |
| 10. Kokoro Text to Speech | 1,064 |
The model in this family
| Model | Use | Privacy | Price |
|---|---|---|---|
Chatterbox HD (Resemble AI)tts-chatterbox-hd | Voice | Private | $0.0225 per 1,000 characters$0.0674 a page read aloud |
Specifications
| Specification | Chatterbox HD |
|---|---|
| Released | May 2025 |
| Weights | Open, MIT |
| Accepts | text, audio |
| Produces | audio |
| Languages | The open model line now covers 23 languages |
| Price | $0.0225 per 1,000 characters |
| Privacy | Private |
Prices, context and lengths from the live catalog on October 16, 2026; parameters and licenses from the maker.
Sources
Everything on this page comes from these pages, read on October 16, 2026 or as dated on each score.