Freedom251 › Voice Cloning Software
MiniMax Speech
What MiniMax Speech is
MiniMax Speech offers text-to-speech, voice cloning, and real-time voice agents through web, API, and self-hosted options. It supports 40 languages, with audio exports in MP3, PCM, FLAC, and WAV. Speech-2.8-HD provides fine-grained emotion and style control and can generate audio up to 30 minutes long; it allows synchronous requests up to 10,000 characters and asynchronous requests up to 1,000,000 characters. Speech-2.8-Turbo is designed for latency-sensitive workloads and supports low-latency PCM streaming, with a 10,000-character synchronous request limit. Instant Voice Clone uses 10 seconds of reference audio and is described as ready in under 30 seconds; it has pay-per-use pricing and no setup fee. Professional Voice Clone uses 30 minutes of curated recordings, takes 3 to 5 business days to train, and is available in cloud or on-premise deployments. Voice agents can handle interruptions, resume with state intact, and switch languages mid-call without swapping models. Listed integrations include LiveKit, Pipecat, custom WebRTC pipelines, and SIP trunks. MiniMax says cloud and on-premise deployment are available. There is no free plan.
Who it suits
It suits developers and teams seeking speech generation, voice cloning, or real-time voice agents through an API, web interface, or self-hosted deployment.
What it does well
- Supports 40 languages.
- Exports MP3, PCM, FLAC, and WAV.
- HD supports asynchronous requests up to 1,000,000 characters.
- Offers cloud and on-premise deployment.
What to know first
- No free plan.
- Professional Voice Clone training takes 3 to 5 business days.
Limits
Speech-2.8-Turbo and Speech-2.8-HD allow up to 10,000 characters per synchronous request. Speech-2.8-HD allows up to 1,000,000 characters asynchronously; Professional Voice Clone requires 30 minutes of recordings and 3 to 5 business days of training.
Plans
| Plan | Billed and limits | Price |
|---|---|---|
| Speech-2.8-HD | $100 per 1M characters · 32kHz audio · 40+ languages · PVC custom voices · Long-form generation up to 30 minutes | $100, paid once |
| Speech-2.8-Turbo | $60 per 1M characters · Sub-200ms time-to-first-byte · PCM streaming · 40+ languages · IVC custom voices | $60, paid once |
As minimax-speech.com lists them, read 8 Oct 2026. The maker's own page has the final word.
Getting started
- Open the MiniMax Speech website
- Choose web, API, or self-hosted deployment
- Select Speech-2.8-Turbo or Speech-2.8-HD
- Provide reference audio for Instant Voice Clone or recordings for Professional Voice Clone
- Connect a voice agent through LiveKit, Pipecat, WebRTC, or SIP
- Use the API reference documentation or book the 30-minute walkthrough
Verdict
MiniMax Speech combines speech generation, voice cloning, and voice agents with several deployment options. Compare the Turbo and HD request limits and the different time requirements for instant and professional cloning.
Questions
Does MiniMax Speech have a free plan?
No. MiniMax Speech is paid and has no free plan.
Which platforms are supported?
It is available through web, API, and self-hosted deployments.
How many languages does it support?
The voice tools support 40 languages.
What audio formats can be exported?
Exports include MP3, PCM, FLAC, and WAV.
What voice-cloning options are available?
Instant Voice Clone uses 10 seconds of reference audio and is described as ready in under 30 seconds. Professional Voice Clone uses 30 minutes of curated recordings and takes 3 to 5 business days to train.
Facts from MiniMax Speech's own pages
- API access
- Yes · minimax-speech.com, 20 Sept 2026
- Deployment
- MiniMax says cloud and on-premise deployment are available. · minimax-speech.com, 8 Oct 2026
- Export formats
- mp3,pcm,flac,wav · minimax-speech.com, 20 Sept 2026
- Free plan
- No · minimax-speech.com, 20 Sept 2026
- HD voice generation
- Speech-2.8-HD supports fine-grained emotion and style control and generates audio up to 30 minutes long. · minimax-speech.com, 8 Oct 2026
- Instant voice cloning
- Instant Voice Clone uses 10 seconds of reference audio and is described as ready in under 30 seconds, with pay-per-use pricing and no setup fee. · minimax-speech.com, 8 Oct 2026
- Integrations
- The voice agent supports LiveKit, Pipecat, custom WebRTC pipelines, and SIP trunks. · minimax-speech.com, 8 Oct 2026
- Languages
- 40 · minimax-speech.com, 20 Sept 2026
- Maker
- MiniMax describes itself as a global AI foundation model company founded in early 2022. · minimax.io, 8 Oct 2026
- Maximum input
- 10000 · minimax-speech.com, 20 Sept 2026
- Platforms
- web,api,self_hosted · minimax-speech.com, 20 Sept 2026
- Professional voice cloning
- Professional Voice Clone uses 30 minutes of curated recordings, takes 3 to 5 business days to train, and is available in cloud or on-premise deployments. · minimax-speech.com, 8 Oct 2026
- Security
- The MiniMax API Privacy Policy says data exchanged with its servers is protected using technologies including HTTPS, SSL, and AES. · platform.minimax.io, 8 Oct 2026
- Support
- The product site offers a 30-minute walkthrough with the solutions team and directs users to API reference documentation. · minimax-speech.com, 8 Oct 2026
- Turbo voice generation
- Speech-2.8-Turbo is designed for latency-sensitive workloads and supports low-latency PCM streaming. · minimax-speech.com, 8 Oct 2026
- Voice agent features
- Voice agents support mid-sentence interruption, resuming with state intact, and switching languages mid-call without swapping models. · minimax-speech.com, 8 Oct 2026
- Voice cloning
- Yes · minimax-speech.com, 20 Sept 2026
- What it does
- MiniMax Voice offers text-to-speech, voice cloning, and real-time voice agents in 40+ languages. · minimax-speech.com, 8 Oct 2026
Where MiniMax Speech stands on Freedom251
- Voice Cloning Software#20 of 25
- Text-to-speech tools#187 of 232
- Text-to-Speech Software#193 of 223