Multilingual speech synthesis with free-form and paralinguistic control.

🤗 Hugging Face｜ 🤖 ModelScope ｜ 📑 Blog & Demos ｜ 💬 DingTalk(钉钉)

Ming-3.0-tts provides two speech-generation models: 0.5b, a dense model, and tiny, a MoE model built on Ling V3 (7.9B total · 1.3B active).

🏆 Artificial Analysis Leaderboards

On the independent Artificial Analysis text-to-speech leaderboards (October 2026), Ming-3.0-tts ranks 1st among open-weights models on Controlled Voice, 1st overall on Pronunciation Robustness, and 3rd among open-weights models on Provider Voice. Ranks change as new models are added.

Within Pronunciation Robustness, Ming-3.0-tts also ranks 1st on Expanding Shorthand (text normalization, 96.5%) and Preserving Exact Sequences (sequence fidelity, 90.6%).

Featured Abilities

Voice creation: clone a reference voice, design one from text, or choose from 100+ Chinese and five English built-in voices.

Free-form control: guide speech delivery with natural-language instructions.

Paralinguistic control: insert 14 inline tags for laughter, sighs, and other vocal expressions in Chinese or English.

Languages and dialects: synthesize Chinese, English, German, Spanish, French, Korean and Japanese, plus Sichuanese and Cantonese.

Emotion and attributes: control emotion, or adjust speed, pitch and volume.

Multi-speaker dialogue: generate a conversation from a script and speaker references (tiny only).

Capabilities use separate task modes, not arbitrary combinations. See the Cookbook for supported inputs, Performance for benchmarks, and Blog & Demos for audio examples.

Architecture

Text and task-specific control tokens condition a language model with a DiT patch head and an audio tokenizer. Both models share this generation framework.

🔥 [2026.10] Ming-3.0-tts-0.5b and Ming-3.0-tts-tiny are released, with inference code, a task cookbook and the evaluation tables below.

Ming-3.0-tts-0.5b weights and inference code

Ming-3.0-tts-tiny weights and inference code

Ming-3.0-tts-tiny benchmarks

vLLM-Omni accelerated inference

Technical report

Quick Start

Built-in voices

Performance

Limitations

Usage Disclaimer

Each release checkpoint bundles the backbone, DiT head, audio tokenizer, speaker encoder and text tokenizer. The 0.5b release layout is shown below; tiny uses sharded weights with model.safetensors.index.json.

Quick Start 🚀

Installation

Or build the 0.5b Docker image (torch 2.6, CUDA 12.4, transformers==4.46.3):

The 0.5b environment has been tested on NVIDIA H800-80GB and H20-96G. For 0.5b, flash-attn is optional; inference falls back to PyTorch SDPA.

Tiny uses a separate reference inference environment: Python 3.10.15, PyTorch and torchaudio 2.7.1+cu118 (CUDA 11.8 wheels), and transformers==4.46.3. Install the pinned dependencies from requirements-tiny.txt in a dedicated environment, as other dependency versions differ from 0.5b. We recommend using NVIDIA H800 for inference.

Model Download

In mainland China, ModelScope is usually faster:

Replace 0.5b with tiny for the other variant.

Text Normalization

Numbers, dates, units and symbols are normalized with FST grammars, which need pynini. It has no universal wheel:

It is loaded lazily; pass --no_text_normalization to skip it.

Ming-3.0-tts-0.5b

Or from Python:

Ming-3.0-tts-tiny

Every task has a one-line script in examples/one_click/; variables such as MODEL, TEXT and OUT can be overridden from the environment:

The same call from Python. Loading configures the release runtime (bf16, portable MoE loop, fp32 recurrent state) and checks the checkpoint against it; generate() defaults to cfg_scale=2.0, temperature=0, flow_steps=10, sigma=0.25, and seed 1234 (1235 for podcast):

Each call selects one task and accepts the inputs listed below:

python infer_tiny.py --help lists the same arguments for the command line; repeat --prompt_wav and --prompt_text for podcast.

Verify Your Setup

Check the checkpoint and tokenizer without loading weights:

Ordered the way a line gets made: who speaks, how it is said, what goes beyond the words, and where it is heard. Each example is one task; controls from another task are rejected. Texts and controls are the ones on the demo page, so you can compare against what you hear there. 0.5b API; full value lists in English / 简体中文, every task end to end in cookbooks/cookbook.ipynb.

Zero-shot Cloning: the TTS call in Quick Start clones the voice of prompt_wav_path.

Voice Design: describe the voice; no reference audio.

Built-in Voices: recommended US_James, US_Stella, 灵小柔 and 灵楷. See the full voice list.

How it is said

Freeform: describe the voice and delivery with one natural-language instruction. Both models default to generation without reference audio; adding a reference is not recommended. Not combined with voice design, a built-in voice, emotion, dialect or attribute controls.

For Tiny, omit prompt_wav and prompt_text:

Emotion: 高兴 / 悲伤 / 愤怒 / 恐惧 / 厌恶 / 惊讶 / 中性.

Speed, Pitch and Volume: speed 慢速 / 中速 / 快速, pitch and volume 低 / 中 / 高.

Beyond the words

Paralinguistic Tags: inline, anywhere in the sentence, in Chinese or English. [Breathing], [Cough], [Sigh], [Laughter], [Surprise-ah], [Surprise-oh], [Dissatisfaction-hnn], [Uhm], [Confirmation-en], [Question-ei], [Question-ah], [Question-en], [Question-oh], [Question-huh].

Where it is heard