text-to-speech

Suno has launched Speech in public beta on web and mobile. It generates spoken voice from a script or text prompt, with optional AI-generated background music, for clips up to roughly eight minutes. Suno claims it is the first audio model to generate voice and music together as one track.

Microsoft AI released MAI-Transcribe-2-Streaming, a real-time transcription model covering 60 languages with first partial results in just over 100 milliseconds. It also launched two text-to-speech models, MAI-Voice-2.1 and MAI-Voice-2.1-Flash, aimed at voice agents.

AWS published a deployment guide for running Alibaba's Qwen3-TTS-12Hz-1.7B-Base voice cloning model as a real-time SageMaker inference endpoint. The model clones a speaker's voice from a short audio clip and generates speech in 10 languages, including cross-lingual cloning, without retraining.

Google DeepMind has released Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, text-to-speech models that generate custom voices from natural language prompts and support line-by-line performance direction. The models top Hume AI's Voice Design Benchmark at 71.4 and claim first and second place on its Overall Quality Index.

Alibaba's Qwen team has released Qwen-Audio-3.1, a five-model lineup covering speech recognition, text-to-speech, and real-time voice interaction. Alongside the release, Alibaba cut API pricing by up to 95 percent for ASR, 85 percent for real-time models, and 70 percent for TTS.

Tencent has open-sourced AuK, a 1.5B-parameter foundation model for speech generation and editing that handles TTS, content editing, and audio enhancement through natural-language instructions. The release includes a distilled AuK-Flash variant for 4-step fast inference, both under MIT license.

NVIDIA's Magpie TTS Multilingual, a 364M-parameter open-weights text-to-speech model, now supports 12 languages after adding Modern Standard Arabic, Korean, and Brazilian Portuguese. The model achieves 32ms time-to-first-audio on B200 GPUs and improves speech quality across French, Spanish, and German.

NVIDIA released Audex-30B-A3B, a unified audio-text model built on the Nemotron-Cascade-2-30B-A3B backbone. The model handles audio understanding, speech recognition and translation, text-to-speech, audio generation, and speech-to-speech while supporting up to 1M token context length.

Mistral AI has released Voxtral TTS, a 4B parameter text-to-speech model supporting 9 languages including English, French, German, Spanish, Dutch, Portuguese, Italian, Hindi, and Arabic. The model achieves 70ms latency for typical inputs and can clone voices from as little as 3 seconds of audio, priced at $0.016 per 1,000 characters.

Supertone has released Supertonic 3, a 99M-parameter text-to-speech model that runs entirely on-device using ONNX Runtime. The model expands language support from 5 to 31 languages compared to Supertonic 2, requires no GPU, and claims competitive accuracy against models 7-20x larger.

Google released Gemini 3.1 Flash TTS, a text-to-speech model that accepts detailed prompts to control voice characteristics, speaking style, accent, and delivery. The model is available through the standard Gemini API using the model ID 'gemini-3.1-flash-tts-preview'.

Google DeepMind launched Gemini 3.1 Flash TTS, a text-to-speech model that achieved an Elo score of 1,211 on the Artificial Analysis TTS leaderboard. The model introduces audio tags that allow developers to control vocal style, pace, and delivery through natural language commands embedded in text input, with support for 70+ languages.

Google's Gemini Live is experiencing persistent voice quality issues following the recent Gemini 3.1 Flash Live update. Users report that voice options like "Capella" (British female accent) have deteriorated significantly, with speech patterns changing dramatically during conversations and audio artifacts like crackles and pops becoming prominent.

Mistral has released Voxtral TTS, a 4-billion-parameter text-to-speech model that can clone voices from just three seconds of reference audio across nine languages. The model delivers 70ms latency for typical 10-second samples and outperformed ElevenLabs Flash v2.5 in naturalness tests. Voxtral is available via API at $0.016 per 1,000 characters and as open-weights on Hugging Face.

Amazon has released a new Bidirectional Streaming API for Amazon Polly that enables simultaneous text input and audio output over a single HTTP/2 connection. The API reduces end-to-end latency by 39% compared to traditional request-response TTS by allowing text to be sent word-by-word as LLMs generate tokens, rather than waiting for complete sentences. The feature is available in Java, JavaScript, .NET, C++, Go, Kotlin, PHP, Ruby, Rust, and Swift SDKs.

Mistral AI released Voxtral-4B-TTS-2603, an open-weights text-to-speech model designed for production voice agents. The 4B-parameter model supports 9 languages, 20 preset voices, achieves 70ms latency at concurrency 1 on a single NVIDIA H200, and requires only 16GB GPU memory.

Mistral AI released Voxtral TTS, an open-source text-to-speech model designed for enterprise voice agents and edge devices. The model supports nine languages, adapts custom voices from samples under five seconds, and achieves 90ms time-to-first-audio latency with a 6x real-time factor.

Xiaomi shipped three AI models simultaneously designed to form a complete agent platform. MiMo-V2-Pro, a 1-trillion-parameter Mixture-of-Experts model with 42 billion active parameters per request, scores 78% on SWE-bench Verified and 81 points on ClawEval—nearly matching Claude Opus 4.6 while costing $1 per million input tokens versus $5 for Opus.

Hume AI has released TADA-1B, a 1 billion parameter text-to-speech model available on Hugging Face under an MIT license. The model, which combines speech and language capabilities, has already accumulated over 3,100 downloads since its January 12 release.