xAI Quietly Ships Speech-to-Text and Text-to-Speech APIs: 25 Languages, Word-Level Timestamps, Real-Time WebSocket Streaming

DEV Community·medium signal

Alongside Grok 4.3 Beta, xAI launched standalone STT and TTS APIs supporting 25+ languages and 12 audio formats. Key features: word-level timestamps, multichannel support, Inverse Text Normalization, speaker diarization, real-time WebSocket streaming, and batch processing via REST. These are production-grade multimodal APIs that compete directly with OpenAI's Whisper API and ElevenLabs, now available through xAI's infrastructure.