e4cc28be8e · 2026-05-03 06:09:26 +07:00
1 Commit
2026-05-03 06:09:26 +07:00
2026-05-03 06:09:26 +07:00
2026-05-03 06:09:26 +07:00
2026-05-03 06:09:26 +07:00
2026-05-03 06:09:26 +07:00
2026-05-03 06:09:26 +07:00
2026-05-03 06:09:26 +07:00
2026-05-03 06:09:26 +07:00

mimotts-cli

Tiny CLI for Xiaomi MiMo TTS.

Install

uv venv
uv pip install -e .

Or run without activating:

uv run mimotts --help

Auth

Create .env:

MIMO_API_KEY=your_key_here
# optional
MIMO_BASE_URL=https://api.xiaomimimo.com/v1

Configure a built-in voice

uv run mimotts config --preset moli

Interactive:

uv run mimotts config

Voice names included here:

  • mimo_default
  • default_zh
  • default_en
  • bingtang
  • moli
  • suda
  • baihua
  • mia
  • chloe
  • milo
  • dean

The Chinese display names are also accepted for the v2.5 voices: 冰糖, 茉莉, 苏打, 白桦.

Configure a designed voice

uv run mimotts config --design "young warm male voice, clear, calm, slight southern accent"

Start a session

uv run mimotts session

Type text, press Enter, get audio. Each generation is saved under workspace/<session-id>/ with a JSONL history file.

Generate once

uv run mimotts say "hello from this cursed chat-completion tts api" --play

Export an old generated file

uv run mimotts export workspace/latest.jsonl --index -1 --to out.wav
S
Description
CHATGPT made this in ONE prompt
Readme
208 KiB
0 Stars 1 Watchers 0 Forks
Languages
Python 100%