MiniMax Speech 2.8 HD

MiniMax Speech 2.8 HD reads your text aloud in 40 languages with 332 voices; emotions, laughs and pauses go right in the text. $0.10 per 1,000 characters.

MiniMax Speech 2.8 HD is the high-quality model of MiniMax's Speech 2.8 family; MiniMax calls it ultra-realistic, with sound tags. It reads a text aloud in one of 332 system voices in 24 languages, 45 of them English, for example English_expressive_narrator, English_Graceful_Lady and English_radiant_girl, and speaks 40 languages and dialects. Short sounds such as (laughs), (sighs) or (breath) and pauses such as <#0.5#> go right into the text. Use it for narration and audiobooks, video voice-overs, podcasts, voice assistants, game characters and read-aloud buttons.

Inputs: text (required: up to 3,000 characters, a Chinese character counting as 2; a longer text is refused, never cut), voice_id (English_expressive_narrator by default: any of MiniMax's system voices, case-sensitive; the list is at platform.minimax.io/docs/faq/system-voice-id), speed (0.5-2, 1 by default), volume (above 0 up to 10, 1 by default), pitch (-12 to 12, 0 by default), emotion (happy, sad, angry, fearful, disgusted, surprised or calm; left empty, the model picks it from the text), language_boost (the language of the text, or auto; empty by default), format (mp3 by default, or wav), sample_rate (8000, 16000, 22050, 24000, 32000 or 44100; 32000 by default) and bitrate (mp3 only: 32000, 64000, 128000 or 256000; 128000 by default).

The result is audio, one MP3 or WAV file (mono), and duration_seconds. It uses MiniMax's built-in voices: it does not clone or design voices, mix voices, make subtitles or take a pronunciation dictionary, and it is not for a text of more than 3,000 characters: send a longer one in parts. In our tests 81 characters took 1.9 s and made 6 to 8 s of speech, and 600 characters took 5.6 s. MiniMax allows 60 requests a minute for the whole account: when it is at its limit a run waits up to a minute and then ends without a charge.

Price: $100 per 1M characters, $0.10 per 1,000, billed by the characters MiniMax counts: a Chinese character (and a Japanese kanji) counts as 2, any other character as 1, with spaces, tags such as (laughs) and pauses such as <#0.5#> as written. The default text (81 characters) costs $0.0081, 1,000 characters $0.10 and the longest text (3,000 characters) $0.30. A refused or failed run costs nothing.

Price: $0.10 per 1,000 characters of text (a Chinese character counts as 2); a request takes up to 3,000 characters, so one costs at most $0.30. One run with default settings costs $0.0081.

Run MiniMax Speech 2.8 HD via API

curl -X POST https://api.flymy.ai/api/v1/flymyai/minimax-speech-2_8-hd/predict \
  -H "X-API-KEY: $FLYMYAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text":"Welcome to FlyMy.AI. Your order has shipped and will arrive on Friday.","voice_id":"English_expressive_narrator"}'

API documentation · Create a free account · Pricing

Related models