Skip to main content
This example demonstrates Fliiq’s translation capability paired with text-to-speech — translate text and get native audio pronunciation delivered straight to Telegram.

The Setup

A user traveling to Hong Kong wanted to translate phrases on the go and hear how they should sound. They told Fliiq:

How It Works

The user already had a Telegram bot set up. Through the real-time Telegram listener, they message Fliiq directly:
Fliiq translates the phrase, generates audio, and replies:
The audio file follows immediately — generated by text_to_speech with the Cantonese_GentleLady voice and sent via send_telegram_audio.

Multi-Language Support

The same pattern works across languages. The agent picks the right TTS voice based on the target language:

English → Mandarin

Cantonese → English

Translation works in both directions:

As a Scheduled Job

For systematic learning, set up a daily translation challenge:

Available Voices

The text_to_speech skill supports multiple voices for different languages and styles: See the full voice list in the Core Skills reference.

The Key Insight

Translation isn’t just text — pronunciation matters. By pairing the LLM’s translation ability with MiniMax text-to-speech and Telegram delivery, Fliiq becomes a pocket translator that you can hear. Send a message, get back a translation with audio you can play to a taxi driver, waiter, or shopkeeper. For languages like Cantonese where tones completely change meaning, hearing the correct pronunciation is essential — not optional.