Voice · MiniMax
MiniMax Speech-02 HD
MiniMax's high-fidelity voice model for voice-overs and audiobooks, in 30+ languages. Pick it for long, steady reading with character voices.
Anonymized Very high costWhat sets it apart
MiniMax Speech-02 HD is an autoregressive transformer with a learnable speaker encoder, aimed at audiobooks and voice-overs in 32 languages. MiniMax has since shipped Speech 2.6 and 2.8, which score higher on the arena.
- 32 languages, zero-shot voice cloning
- HD and Turbo variants
- Later 2.8 adds natural sound tags
Strengths and limits
Where it shines
- More than 30 languages, French included
- 15 character voices (calm, deep, lively...)
Think twice when
- Among the dearest voices per character
- Voice cloning is not offered in Sub Rosa
- 15 voices
- Up to 5,000 characters per request in Sub Rosa
Use cases
Audiobook chapter
Read this chapter in a deep, calm male voice, steady pace.
Steady long reading with character voices.
Corporate voice-over
Narrate this 45 second company intro in a confident, friendly tone.
HD quality in 32 languages.
Against its rivals
- ElevenLabsElevenLabs is about half the price at a similar arena score.#1 of 10 on Artificial Analysis Speech Arena (1,330 Elo)
- Gemini 3.1 Flash TTSGemini scores higher and uses tags, at a higher price.#3 of 10 on Artificial Analysis Speech Arena (1,209 Elo)
- Gradium TTSGradium scores higher at about a third of the price.#5 of 10 on Artificial Analysis Speech Arena (1,151 Elo)
Version history
Each version, dated, with what it changed over the one before.
- Speech-02 HDLatest
- Introduced HD and Turbo variants with strong prosody and voice similarity
- Speech 2.6Earlier, not in Sub Rosa
- More responsive voice for agents
- Speech 2.8Earlier, not in Sub Rosa
- Natural sound tags and cleaner audio
Benchmarks
How to read these scores: what a benchmark tells you.
Artificial Analysis Speech Arena: #6 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 |
|---|---|---|---|
MiniMax Speech-02 HDtts-minimax-speech-02-hd | Voice | Anonymized | $0.0562 per 1,000 characters$0.169 a page read aloud |
Specifications
| Specification | Speech-02 HD |
|---|---|
| Released | April 2, 2025 |
| Weights | Closed |
| Accepts | text |
| Produces | audio |
| Languages | 32 languages |
| Price | $0.0562 per 1,000 characters |
| Privacy | Anonymized |
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.