GPT-Realtime-1.5

Grok Voice Think Fast 2.0 is xAI’s flagship voice model for building real-time assistants, phone agents, and interactive voice systems that stream audio and text bidirectionally over WebSocket. Developers can configure system instructions, high or no reasoning effort, built-in or custom voices, automatic server-side voice activity detection, silence duration, idle re-engagement, playback speed, and session resumption after temporary disconnects. It accepts PCM, G.711 μ-law, G.711 A-law, and Opus audio through JSON or raw binary frames, with configurable PCM sample rates from telephone quality to 48 kHz. It supports more than 20 languages with native-quality accents, automatic language detection, natural responses in the speaker’s language, and seamless code-switching. Language hints and up to 100 key terms improve transcription of regional speech, names, products, codes, addresses, and specialized terminology, while pronunciation replacements correct spoken output.

Voxtral Mini TTS 3.0 is Mistral AI's text-to-speech model for developers building voice-enabled applications and services. It generates speech from text in batch and real-time workflows, with preset or cloned voices, emotion and pace controls, and custom pronunciation. The model supports multilingual speech generation and dialect adaptation, making it suitable for conversational audio, localized content, spoken responses, and voice agents. Developers can access speech generation and voice-cloning capabilities through Mistral's API. Voxtral Mini TTS 3.0 is also part of Mistral's Voxtral 3 voice-agent stack, which combines speech recognition, reasoning, tool use, and generated speech; the Voxtral 3 stack can be self-deployed on customer infrastructure.