Kokoro vs Chatterbox Turbo
Start with Kokoro for compact preset-voice narration. Try Chatterbox Turbo for a more expressive delivery or a workflow that needs reference-voice conditioning.
Editorial guidance · Reviewed September 14, 2026 · No universal winner
When to choose Kokoro-82M
Kokoro keeps the workflow small: choose a preset voice and synthesize speech, including on CPU. It is well suited to a straightforward script-to-audio tool.
Samples, sources & setup →When to choose Chatterbox Turbo
Chatterbox Turbo is a newer release in the Chatterbox family. Its expressive controls and voice-cloning workflow offer a different set of tradeoffs; the recordings here are the Turbo checkpoint, not the original Chatterbox model.
Samples, sources & setup →Hear the same three scripts
Use the emotional script to compare phrasing, then check the numbers script for pronunciation. Both models accept your own English text as separately quoted paid recordings with different length limits.
Neutral
“The quick brown fox jumps over the lazy dog near the riverbank.”
Bella
Andy
Emotional
“I can't believe you actually did it. This is incredible!”
Bella
Andy
Numbers & Dates
“On March 14th, 2025, the team raised $4.2 million at a 38% margin.”
Bella
Andy
Try your text with a hosted preset
Open a model below, paste your script, and review its individual quote. Each model records separately using prepaid credits from $5. To compare your own text, create a separate recording with each available model; there is no automatic side-by-side generation.
The hosted studio exposes English preset voices, with model-specific limits. Directory recordings may use other voices or upstream features. The existing samples here remain free to hear.
Kokoro-82M
Eight English voices. Edit and quote each paragraph, then export one MP3 or WAV.
Create a paid Kokoro voiceover →Chatterbox Turbo
Andy preset · Up to 500 characters per recording. Review this model’s quote before generating.
Record with Chatterbox Turbo (paid) →Practical differences
| Feature | Kokoro-82M | Chatterbox Turbo |
|---|---|---|
| License | Apache-2.0 | MIT |
| Parameters | 82M | 350M |
| Voice cloning | Preset voices | Supported by the model |
| Languages | English, Japanese, Chinese, Spanish, French, Hindi, Italian, Portuguese | English |
| This recording | Bella | Chatterbox Turbo · Replicate preset Andy |
Code and weight licenses may differ. Follow the primary sources on each model page. We do not publish a speed ranking without a controlled benchmark.