gemini-tts
Installation
SKILL.md
Gemini TTS - Read-Aloud Speech Skill
Use the Python script in scripts/ to turn text into natural read-aloud audio
via Google Gemini TTS. Model: gemini-3.1-flash-tts-preview (single TTS
model — there is no Pro variant).
The mode is automatically detected from the input:
| Mode | When used | Voices |
|---|---|---|
| Single-speaker | Plain text / narration | One voice (--voice) |
| Multi-speaker | A 2-person dialogue: lines like Name: ... with exactly two distinct speakers |
Two voices (--speaker) |
- Detected 2 speaker labels → multi-speaker (2 is a hard limit)
- Detected 1 or 0 → single-speaker
- Detected 3+ → warns and falls back to single-speaker narration
- Override with
--single/--multi