No official torrent yet — requesting it tells the flock what to host next. You cannot seed a catalog-only page.
Cataloged from OpenMOSS-Team/MOSS-Transcribe-Diarize
All models by OpenMOSS-Team →
MOSS-Transcribe-Diarize 0.9B is an end-to-end audio understanding model for long-form multi-speaker transcription, diarization, timestamps, and acoustic event awareness.
It supports transcription and diarization across 50+ languages, single-pass inference on audio recordings up to 90 minutes long, and custom hotword prompting for domain-specific terms.
Given an audio or video file, the model generates a compact speaker-aware transcript in one pass, including timestamps and anonymous speaker labels such as [S01], [S02], and beyond.
MOSS-Transcribe-Diarize 0.9B turns real-world long-form audio into structured, speaker-aware transcripts in one pass. Instead of stitching together separate ASR and diarization systems, it jointly performs speech transcription and speaker diarization, producing time-aligned text with consistent speaker labels.
The model is built for meetings, calls, podcasts, interviews, lectures, videos, and other long or messy multi-speaker recordings. It can also emit acoustic event annotations, giving downstream systems a richer view of what happened, who spoke, and when.
Core capabilities:
[S01] and [S02] without a separate diarization pipeline.[Image blocked: MOSS-Transcribe-Diarize 0.9B model architecture]
This Hugging Face repository includes the custom Transformers remote code required to load the model with trust_remote_code=True.
We evaluate MOSS-Transcribe-Diarize using three objective metrics: Character Error Rate (CER), concatenated minimum-permutation Character Error Rate (cpCER), and Delta-cp. Lower is better for all metrics. A dash (-) indicates that the result is unavailable.
Use a clean Python environment. The model uses custom Transformers code, so load the model and processor with trust_remote_code=True.
The GitHub package provides helper utilities such as audio/video loading, transcription message construction, transcript parsing, CLI inference, and the subtitle web app. The model weights and remote-code model files are loaded from this Hugging Face repository.
The message flow follows the common Qwen multimodal pattern:
processor.apply_chat_template(messages, tokenize=False) renders text with audio placeholders.processor(text=text, audio=audios) computes Whisper input features and expands audio placeholders.model.generate(...) produces timestamped transcription and diarization text.The default prompt is optimized for timestamped transcription and speaker diarization:
To add hotwords, append a short hint to the default prompt:
More prompt recipes are available in the GitHub repository: https://github.com/OpenMOSS/MOSS-Transcribe-Diarize/blob/main/examples/prompts.md
The recommended way to serve MOSS-Transcribe-Diarize is SGLang Omni through the OpenAI-compatible /v1/audio/transcriptions endpoint. If you are using a CUDA 12 environment, SGLang is currently not supported; use vLLM instead. Install sglang-omni by following the installation guide, then download the model:
Serve the model:
Use response_format=verbose_json when you need parsed speaker segments. json returns the raw transcript text only.
For longer multi-speaker audio, raise max_new_tokens so the decoder can finish the full diarized transcript:
MOSS-Transcribe-Diarize also supports vLLM serving through the OpenAI-compatible transcription API. Use a pinned vLLM nightly build that includes the MOSS-Transcribe-Diarize model registration. Choose one of the following commands: for CUDA 12 environments, use cu129; for CUDA 13 environments, use cu130.
or:
The source package includes a local subtitle workflow for upload, review, subtitle export, and optional FFmpeg burn-in:
Open http://127.0.0.1:7860, upload an audio/video file, review the parsed subtitle segments, then download JSON/SRT/ASS or burn an MP4 if ffmpeg and ffprobe are available on PATH.
For batch processing:
The canonical output format is:
Example:
In this format:
start_time and end_time are timestamps in seconds.[S01], [S02], and similar labels are anonymous model-generated speaker labels.MOSS-Transcribe-Diarize 0.9B is licensed under the Apache License 2.0.
If you use MOSS-Transcribe-Diarize 0.9B, please cite the technical report:
conda create -n moss-transcribe-diarize python=3.12 -yconda activate moss-transcribe-diarize
git clone https://github.com/OpenMOSS/MOSS-Transcribe-Diarize.gitcd MOSS-Transcribe-Diarize
pip install --index-url https://download.pytorch.org/whl/cu128 torch torchaudiopip install -e .import torchfrom transformers import AutoModelForCausalLM, AutoProcessor
from moss_transcribe_diarize import parse_transcriptfrom moss_transcribe_diarize.inference_utils import ( build_transcription_messages, generate_transcription, resolve_device,)
model_id = "OpenMOSS-Team/MOSS-Transcribe-Diarize"audio_path = "audio.wav"
device = resolve_device("auto")dtype = torch.bfloat16 if device.type == "cuda" else torch.float32
model = AutoModelForCausalLM.from_pretrained( model_id, trust_remote_code=True, dtype="auto",).to(dtype=dtype).to(device).eval()
processor = AutoProcessor.from_pretrained( model_id, trust_remote_code=True,)
messages = build_transcription_messages(audio_path)result = generate_transcription( model, processor, messages, max_new_tokens=2048, do_sample=False, device=device, dtype=dtype,)
print(result["text"])
for segment in parse_transcript(result["text"]): print(segment.start, segment.end, segment.speaker, segment.text)请将音频转写为文本，每一段需以起始时间戳和说话人编号（[S01]、[S02]、[S03]…）开头，正文为对应的语音内容，并在段末标注结束时间戳，以清晰标明该段语音范围。请将音频转写为文本，每一段需以起始时间戳和说话人编号（[S01]、[S02]、[S03]…）开头，正文为对应的语音内容，并在段末标注结束时间戳，以清晰标明该段语音范围。热词提示：热词1, 热词2, 热词3hf download OpenMOSS-Team/MOSS-Transcribe-Diarizesgl-omni serve \ --model-path OpenMOSS-Team/MOSS-Transcribe-Diarize \ --port 8000 \ --max-running-requests 16 \ --cuda-graph-max-bs 16 \ --mem-fraction-static 0.80curl -X POST http://localhost:8000/v1/audio/transcriptions \ -F model=OpenMOSS-Team/MOSS-Transcribe-Diarize \ -F file=@audio.wav \ -F response_format=verbose_jsoncurl -X POST http://localhost:8000/v1/audio/transcriptions \ -F model=OpenMOSS-Team/MOSS-Transcribe-Diarize \ -F file=@audio.wav \ -F response_format=verbose_json \ -F max_new_tokens=65536uv pip install -U vllm \ --torch-backend=auto \ --extra-index-url https://wheels.vllm.ai/68b4a1d582818e67adc903bf1b8fc5a5447da2fa/cu129uv pip install -U vllm \ --torch-backend=auto \ --extra-index-url https://wheels.vllm.ai/68b4a1d582818e67adc903bf1b8fc5a5447da2fa/cu130vllm serve OpenMOSS-Team/MOSS-Transcribe-Diarize --trust-remote-codecurl http://localhost:8000/v1/audio/transcriptions \ -F model="OpenMOSS-Team/MOSS-Transcribe-Diarize" \ -F file=@"audio.wav" \ -F response_format="json" \ -F temperature="0"mtd-subtitle-web \ --model OpenMOSS-Team/MOSS-Transcribe-Diarize \ --host 127.0.0.1 \ --port 7860mtd-subtitle /path/to/input.mp4 \ --model OpenMOSS-Team/MOSS-Transcribe-Diarize \ --out-dir runs/example \ --render[start_time][Sxx]transcribed speech[end_time][0.48][S01]Welcome everyone[1.66][12.26][S02]The new transcription pipeline is ready for evaluation[13.81][14.36][S01]Great, include the diarization results in the report[18.76]@misc{moss_transcribe_diarize_2026, title={MOSS Transcribe Diarize Technical Report}, author={{MOSI.AI}}, year={2026}, eprint={2601.01554}, archivePrefix={arXiv}, primaryClass={cs.SD}, url={https://arxiv.org/abs/2601.01554}}