Take a slow breath in, hold it for a moment, and let your shoulders drop as you breathe out.
MiniMax Speech HD AI Voice Generation
One voice that keeps its sound across languages.
Affordable text-to-speech and voice cloning in 29 languages.
Bienvenidos a nuestra panadería. El pan sale del horno cada mañana a las siete, y hoy tenemos pan de centeno recién hecho.
- Version
- HD
- Released
- 2025-02-19
- Starts at
- 34 credits
- Run time
- 16 smedian of 2 runs on Oakgen
The same female voice welcoming customers to a bakery in Spanish
A calm female voice guiding a short breathing exercise in English
Steal a prompt. Make it yours.
Every one of these recordings was made on Oakgen with this model.
Bienvenidos a nuestra panadería. El pan sale del horno cada mañana a las siete, y hoy tenemos pan de centeno recién hecho.
Two things MiniMax Speech HD does well.
Same voice, another language.
MiniMax Speech HD speaks 29 languages and keeps the voice's timbre when it switches. Both runs use the same voice, one in English and one in Spanish.
Try another languageA steady read for guided audio.
Paste a script and use punctuation for natural pauses. Output comes back as 44.1 kHz audio.
Make a narrationSpecs
- Starting price
- $0.13 / generation
- Generation time
- 2–5 seconds
- Max resolution
- 44.1 kHz stereo
- Inputs → outputs
- text → audio
List price
34credits to start
Compare
| Model | Starts at · credits |
|---|---|
| MiniMax Speech HDThis pageMiniMax | |
| ElevenLabs Eleven v4ElevenLabs | — |
| ElevenLabs v3ElevenLabs | — |
What is MiniMax Speech HD?
Affordable text-to-speech and voice cloning in 29 languages.
MiniMax Speech HD is MiniMax's high-definition text-to-speech model, released in February 2025. It supports 29 languages and voice cloning from a 30-second sample. On Oakgen it is charged from a $0.10 per 1,000 characters provider cost, the same rate as ElevenLabs v3, with a minimum charge per request.
Capabilities at a glance
- 29 languages with preserved timbre across languages
- 30-second voice cloning from a clean sample
- 44.1 kHz MP3/WAV output
- 2–5 second latency per request
How to use MiniMax Speech HD
Pick a stock voice or clone your own
Browse 150+ stock voices by language, age, and use case. For a cloned voice, upload a 30-second clean sample.
Paste your script
Up to 50,000 characters per request — about 45 minutes of audio. Use punctuation for natural pauses.
Generate
Typical 1-minute script finishes in 3–5 seconds. Download MP3 or WAV.
API access
curl -X POST https://api.oakgen.ai/v1/generate/speech \
-H "Authorization: Bearer $OAKGEN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "minimax-speech-hd",
"voice": "female_alloy",
"text": "Welcome to Oakgen."
}'Compared to other models
- vs. elevenlabs-v3 — cost
- On Oakgen, MiniMax Speech HD and ElevenLabs v3 are charged from the same $0.10 per 1,000 characters provider cost, and MiniMax adds a minimum charge per request. Eleven v4 ($0.08) and Eleven v4 Turbo ($0.04 per 1,000 characters) cost less per character.
License & commercial use
Licensed through MiniMax's commercial terms.
Permitted on all paid Oakgen plans.
Frequently asked questions
How much does MiniMax Speech HD cost on Oakgen?
MiniMax Speech HD starts at $0.13 per generation on Oakgen. Most generations complete in 2–5 seconds.
Can I use MiniMax Speech HD commercially?
Permitted on all paid Oakgen plans.
What is the maximum output resolution?
MiniMax Speech HD supports up to 44.1 kHz stereo.
Does Oakgen provide API access to MiniMax Speech HD?
Yes. Oakgen's REST API exposes MiniMax Speech HD under the model slug 'minimax-speech-hd'. See the API snippet below for an example request.