What problem does it solve? Converting written content into natural-sounding spoken audio requires a TTS service with quality voices, language coverage, and tuning controls, which is cumbersome to integrate manually. This Skill wraps the ElevenLabs text-to-speech API behind the inference.sh CLI so you can generate speech with a single command. ## Core Features & Use Cases - 22+ Voices and 3 Models: Choose from male and female voices across models like eleven_multilingual_v2 (highest quality), eleven_turbo_v2_5, and eleven_flash_v2_5 (low latency). - Voice Tuning and Formats: Adjust stability, similarity_boost, and style, and export as MP3 or raw PCM at multiple sample rates. - Multilingual Support: Generate speech in 32 languages for voiceovers, audiobooks, podcasts, e-learning, IVR, and accessibility content. - Use Case: Produce a Spanish product-demo voiceover with the aria voice, then feed the resulting audio URL into a talking-head video generator. ## Quick Start Ask the assistant to generate speech from the text "Welcome to our product demo" using the ElevenLabs aria voice via the belt CLI.
