Sophi Dzeberashvili
Mingrelian · Standard · Young Adult · female
Calm gentle authoritative feminine
Adapted for peer support conversation
conversational, strong, expressive, energetic, balanced
Voice library / Language
Listen to the top Mingrelian AI voices in Giggy's library and use any of them for text to speech. Batch generation is free, within request and queue limits.
Mingrelian · Standard · Young Adult · female
Calm gentle authoritative feminine
Adapted for peer support conversation
conversational, strong, expressive, energetic, balanced
Mingrelian · Standard · Young Adult · female
Warm gentle affirmation voice
A useful voice for wellness affirmation narration
bright, urgent, raspy, quick, friendly
Mingrelian · Standard · Adult · female
Gentle authoritative narrative presence
A natural choice for biography narration
full, rich, upbeat, serious, friendly
Play samples from 3 Mingrelian voices and choose the one that fits your script, starting with top-rated voices like Sophi Dzeberashvili, Elene Abashidze, Tamar Zhuraveli.
Type or paste your text in Mingrelian. Add expression tags such as [whisper] or [laughing] to guide emotion and delivery.
Batch generation is free, within request and queue limits, and returns 24 kHz MP3 audio. Fast and Streaming use prepaid credits when you need speed.
Most common accents among the top-rated Mingrelian voices.
Every library voice has a voice ID for the Giggy Speech API, its OpenAI-compatible speech endpoint and the Giggy MCP server.
Giggy's voice library has 3 Mingrelian voices. This page shows the highest-rated voices by quality score; the full library is searchable in the Giggy app.
Yes. Batch generation is free in the Giggy app and through the API, within request and queue limits. Fast and Streaming use prepaid credits; there is no subscription.
The most common accents among the top-rated Mingrelian voices are Standard. Use the accent filter in the Voice Library to see every option.
Yes. Voice Design creates a new Mingrelian voice from a written description. Voice Clone reference audio supports a separate set of languages, listed on the voice cloning page.
Yes. Giggy supports 50+ emotions, 20+ delivery styles and 40+ non-verbal events through expression tags, plus speaking-speed control.
Yes. Every library voice has a voice ID you can use with the Giggy Speech API, its OpenAI-compatible speech endpoint and the Giggy MCP server.