About this model
High-quality text-to-speech with expressive, natural voice synthesis. MiniMax Speech 2.8 is an advanced text-to-speech model that turns text into natural, expressive audio in multiple languages.
Pricing
| Rate | RODI | USD (ref.) | Unit |
|---|---|---|---|
| Per unit | |||
| Per 1K characters | 75.8 | ~ 0.1 | per 1K characters (RODI/1K chars, USD/1K chars) |
RODI prices include RodiumAi markup and upstream fees. USD figures are wholesale reference rates.
Providers & pricing
| Provider | Input | Output | Cached | Free tier | Status |
|---|---|---|---|---|---|
 Runware Default All rates (1)
| - | - | - | - | No data |
Capabilities
Streaming
Tool calling
Vision
JSON mode
Reasoning
API usage
Synthesize speech with client.speech(): binary POST /v1/audio/speech.
Examples
…RodiumAi SDK
speech() returns raw audio bytes (save to disk or stream to clients).
…Use cases
- Natural voice-overs and narration in many languages.
- Voice agents, IVR and assistants.
- Audio versions of articles, courses and notifications.
- Character voices for games and videos (the library holds hundreds of voices).
Notes
voiceis one of MiniMax's library voice ids, for exampleEnglish_expressive_narrator(default),English_radiant_girl,English_magnetic_voiced_man,English_Upbeat_Woman. The library also has Chinese, Japanese, Korean and other languages. An OpenAI voice name such asalloyis replaced by the default voice.inputaccepts up to 50 000 characters.- Choose the language with
runware.settings.languageBoost(en,fr,es,pt,ar,zh, ...). The defaultautodetects it from the text.runware.settings.turbo: truetrades some quality for faster, cheaper audio. - Do not send
speedorlanguageat the top level: this model only reads the text and the voice. response_format:mp3(default),wav,flacoropus(delivered as OGG).