MCP / Models / Voice, music and sound

MiniMax Speech 2.8 HD MCP: 332 voices, emotions and laughs

MiniMax Speech 2.8 HD reads text aloud in 40 languages with 332 voices, and takes emotions, laughs and pauses right in the text. Run it from Claude Code, Codex, Cursor or any MCP client through one FlyMy.AI MCP, or call its API.

No API key$0.10 / 1,000 charactersText to speechMiniMaxCommercial use
MiniMax Speech 2.8 HD sample from the FlyMy.AI catalog

$ YOUR NEXT RUN

Claude CodeCodexCursorAny agent

Use MiniMax Speech 2.8 HD in your assistant

Paste this prompt into Claude Code, Codex, Cursor or any agent. It reads the guide and helps you connect.

Read https://flymy.ai/skill.md and https://flymy.ai/mcp/models/minimax-speech-2-8-hd.md. Help me connect the FlyMy.AI MCP to this assistant, then run MiniMax Speech 2.8 HD (flymyai/minimax-speech-2_8-hd) for this task: Read this paragraph aloud with MiniMax Speech 2.8 HD in a calm, friendly voice and send me the audio file.
Open FlyMy.AIDownload brief

What MiniMax Speech 2.8 HD does

Send text, up to 3,000 characters with a Chinese character counting as 2; a longer text is refused, never cut, so send it in parts. Short sounds go right into it as tags such as (laughs), (sighs) or (breath), and a pause as <#0.5#>, in seconds. voice_id is one of MiniMax's 332 system voices in 24 languages (English_expressive_narrator is the default, and the id is case-sensitive). emotion is happy, sad, angry, fearful, disgusted, surprised or calm, and left empty the model picks it from the text. language_boost names the language of the text or lets MiniMax find it; speed, volume and pitch adjust the delivery, and format is mp3 or wav, with sample_rate and bitrate for the file. The answer is the audio and duration_seconds. HD is the high-quality model of MiniMax's Speech 2.8 family. It does not clone or design voices, mix voices or make subtitles. MiniMax allows 60 requests a minute for the whole account, so a run in a busy minute can wait, or end with no charge.

MiniMax Speech 2.8 HD inputs

Every input of flymyai/minimax-speech-2_8-hd as its live schema lists them. Returns audio (array), duration_seconds (number).

InputType, values and default
# text
string required The words to speak, up to 3,000 characters (a Chinese character counts as 2), about 3 minutes of English. A longer text is refused, never cut: send it in parts. Sounds go in as tags, (laughs) (chuckle) (coughs) (clear-throat) (groans) (breath) (pant) (inhale) (exhale) (gasps) (sniffs) (sighs) (snorts) (burps) (lip-smacking) (humming) (hissing) (emm) (sneezes), and a pause as <#0.5#> (seconds, 0.01-99.99, never two in a row); tags and pauses count as characters. Billed $0.10 per 1,000 characters
# voice_id
string optional Default: English_expressive_narrator The voice: any of MiniMax's 332 system voices in 24 languages (45 in English); the id is case-sensitive. English: English_expressive_narrator (default), English_Graceful_Lady, English_radiant_girl, English_magnetic_voiced_man, English_compelling_lady1, English_CaptivatingStoryteller, English_ManWithDeepVoice, English_Aussie_Bloke. Other languages: Spanish_SereneWoman, French_Male_Speech_New, German_FriendlyMan, Italian_Narrator, Portuguese_SentimentalLady, Russian_HandsomeChildhoodFriend, Japanese_IntellectualSenior, Korean_AirheadedGirl, Chinese (Mandarin)_Reliable_Executive, Arabic_CalmWoman, hindi_male_1_v2. The full list is at platform.minimax.io/docs/faq/system-voice-id
# speed
number optional Allowed: 0.5 to 2 Default: 1.0 How fast: 0.5 (half speed) to 2 (double), 1 is normal
# volume
number optional Allowed: 0.1 to 10 Default: 1.0 How loud: above 0 up to 10 (MiniMax's vol), 1 is normal
# pitch
integer optional Allowed: -12 to 12 Default: 0 Pitch shift: -12 (lower) to 12 (higher), 0 is unchanged
# emotion
string optional Allowed: happy, sad, angry, fearful, disgusted, surprised, calm The emotion of the voice: happy, sad, angry, fearful, disgusted, surprised or calm. Leave empty and the model picks it from the text
# language_boost
string optional Allowed: Chinese, Chinese,Yue, English, Arabic, Russian, Spanish, French, Portuguese, German, Turkish, Dutch, Ukrainian, Vietnamese, Indonesian, Japanese, Italian, Korean, Thai, Polish, Romanian, Greek, Czech, Finnish, Hindi, Bulgarian, Danish, Hebrew, Malay, Persian, Slovak, Swedish, Croatian, Filipino, Hungarian, Norwegian, Slovenian, Catalan, Nynorsk, Tamil, Afrikaans, auto The language of the text, to boost its recognition when the model should not guess it: one of MiniMax's 40 languages and dialects (a Cantonese voice needs Chinese,Yue), or auto to let it find the language. Leave empty for none
# format
string optional Allowed: mp3, wav Default: mp3 File type: mp3 is small, wav is 16-bit samples without loss. The price is the same
# sample_rate
integer optional Allowed: 8000 to 44100 Default: 32000 Samples a second: 8000, 16000, 22050, 24000, 32000 or 44100. The price is the same
# bitrate
integer optional Allowed: 32000 to 256000 Default: 128000 mp3 only (a wav ignores it): 32000, 64000, 128000 or 256000. The price is the same

Ask your agent

Requests to paste as they are, once the FlyMy.AI MCP is connected.

“Read this paragraph aloud with MiniMax Speech 2.8 HD in a calm, friendly voice and send me the audio file.”runs flymyai/minimax-speech-2_8-hd
“Voice my product demo script with MiniMax Speech 2.8 HD and give me the audio file.”runs flymyai/minimax-speech-2_8-hd
“With MiniMax Speech 2.8 HD, narrate my video script slowly so I can lay the audio over the footage.”runs flymyai/minimax-speech-2_8-hd

MiniMax Speech 2.8 HD 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. A run at the default settings costs $0.0081. A run costs $0.0001 to $0.30, depending on text. Billed to your FlyMy.AI balance; no seats, no subscription and no minimum.

SettingPrice
Any$0.10 per 1,000 characters

Call MiniMax Speech 2.8 HD from code

The same model, from an agent or from your own code. Keys stay in your FlyMy.AI account.

From an agent, through the MCP

The agent calls get_model_schema, then run_model with:

{
  "endpoint_id": "flymyai/minimax-speech-2_8-hd",
  "input": {
    "text": "Welcome to FlyMy.AI."
  }
}

From your code, over HTTP

Form fields, not JSON; files go as uploads or public URLs. The answer is a prediction_id; read the result at https://api.flymy.ai/api/v1/flymyai/minimax-speech-2_8-hd/predict/async/result/?request_id=<prediction_id>.

curl -X POST https://api.flymy.ai/api/v1/flymyai/minimax-speech-2_8-hd/predict/async/ \
  -H "X-API-KEY: $FLYMYAI_API_KEY" \
  -F 'text=Welcome to FlyMy.AI.'

Questions about MiniMax Speech 2.8 HD

Do I need a MiniMax API key for MiniMax Speech 2.8 HD?

No. MiniMax Speech 2.8 HD runs on FlyMy.AI's account, so you need no MiniMax key or plan: sign in to FlyMy.AI once and each run is billed to your FlyMy.AI balance.

How much does MiniMax Speech 2.8 HD cost?

$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. There is no subscription or minimum; the price comes from FlyMy.AI billing and was checked on 2026-11-18.

How do I use MiniMax Speech 2.8 HD in Claude Code, Codex or Cursor?

Add the FlyMy.AI MCP with one command and sign in, then ask in plain words, for example: "Read this paragraph aloud with MiniMax Speech 2.8 HD in a calm, friendly voice and send me the audio file." Your agent reads the inputs with get_model_schema and runs flymyai/minimax-speech-2_8-hd with run_model.

What does MiniMax Speech 2.8 HD need as input?

MiniMax Speech 2.8 HD requires text; it also takes 9 optional inputs, each listed on this page with its allowed values and default.

Can I call MiniMax Speech 2.8 HD from my own code?

Yes. Send form fields to https://api.flymy.ai/api/v1/flymyai/minimax-speech-2_8-hd/predict/async/ with your FlyMy.AI API key in the X-API-KEY header.

Go to the source

From the FlyMy.AI model catalog of 2026-11-18: inputs from the model's live schema, prices from FlyMy.AI billing. Requests are examples to paste, not recorded runs.