{
  "hf_id": "OpenMOSS-Team/MOSS-TTSD-v1.0",
  "meta": {
    "title": "MOSS-TTSD-v1.0",
    "slug": "moss-ttsd-v1.0",
    "provider": "OpenMOSS",
    "description": "OpenMOSS's 8B spoken-dialogue generation model for expressive, multi-speaker, ultra-long conversations — served via vLLM-Omni through the OpenAI /v1/audio/speech API (24 kHz mono).",
    "date_added": "2026-06-17",
    "date_updated": "2026-06-15",
    "tasks": [
      "omni"
    ],
    "related_recipes": [
      "OpenMOSS-Team/MOSS-TTS",
      "OpenMOSS-Team/MOSS-TTS-Realtime",
      "OpenMOSS-Team/MOSS-VoiceGenerator"
    ]
  },
  "model": {
    "model_id": "OpenMOSS-Team/MOSS-TTSD-v1.0",
    "min_vllm_version": "0.22.0",
    "architecture": "dense",
    "parameter_count": "8B",
    "active_parameters": "8B",
    "context_length": 0,
    "base_args": [
      "--omni",
      "--deploy-config",
      "vllm_omni/deploy/moss_ttsd.yaml"
    ],
    "base_env": {},
    "install": {
      "pip": {
        "command": "uv venv\nsource .venv/bin/activate\nuv pip install -U vllm --torch-backend auto"
      },
      "docker": {
        "command": "docker pull vllm/vllm-openai:latest"
      },
      "extras": [
        {
          "note": "vLLM-Omni provides the TTS serving backend and the `--omni` serve path. See the installation guide for pre-built wheels or building from source.",
          "command": "uv pip install git+https://github.com/vllm-project/vllm-omni.git"
        }
      ]
    }
  },
  "omni": {
    "tasks": [
      {
        "id": "t2a",
        "label": "Text → Speech",
        "endpoint": "/v1/audio/speech",
        "description": "MOSS-TTSD multi-speaker dialogue synthesis via /v1/audio/speech (24 kHz mono)",
        "curl": "curl -X POST http://localhost:8000/v1/audio/speech \\\n  -H \"Content-Type: application/json\" \\\n  -d '{\n    \"model\": \"OpenMOSS-Team/MOSS-TTSD-v1.0\",\n    \"input\": \"[S1] Hi, how was your day? [S2] Pretty good, thanks for asking!\",\n    \"voice\": \"default\",\n    \"response_format\": \"wav\"\n  }' --output dialogue.wav\n"
      }
    ]
  },
  "features": {},
  "opt_in_features": [],
  "variants": {
    "default": {
      "precision": "bf16",
      "vram_minimum_gb": 32,
      "description": "MOSS-TTSD-v1.0 8B (MossTTSDelay, n_vq=16) — multi-speaker dialogue TTS"
    }
  },
  "hardware_overrides": {},
  "guide": "## Overview\n\n[MOSS-TTSD-v1.0](https://huggingface.co/OpenMOSS-Team/MOSS-TTSD-v1.0) is the\n**spoken-dialogue** member of OpenMOSS's MOSS-TTS Family, served through\n**vLLM-Omni** with the OpenAI-compatible `/v1/audio/speech` API. It generates\nexpressive, **multi-speaker, ultra-long dialogues** and (per OpenMOSS) leads on\nobjective metrics while outperforming top closed-source systems in subjective\nevaluations. Output is **24 kHz mono**.\n\nDialogue formatting (speaker turns, e.g. `[S1] ... [S2] ...`) and any\nmulti-speaker reference conditioning follow the upstream\n[MOSS-TTSD](https://github.com/OpenMOSS/MOSS-TTSD) conventions — consult the\nupstream repo for the exact turn/reference schema.\n\n## Prerequisites\n\n- **Hardware**: a single CUDA GPU comparable to the 8B MOSS-TTS profile\n  (~18 GB talker + ~8 GB codec; e.g. an 80 GB H100 with headroom for long\n  dialogues).\n- **vLLM-Omni** targeting vLLM >= 0.22.\n\n## Installation\n\n```bash\nuv venv && source .venv/bin/activate\nuv pip install git+https://github.com/vllm-project/vllm-omni.git\n```\n\n## Launch the server\n\nMOSS-TTSD shares `model_type=moss_tts_delay` with the other MossTTSDelay\ncheckpoints, so pass its deploy config explicitly:\n\n```bash\nvllm serve OpenMOSS-Team/MOSS-TTSD-v1.0 --omni \\\n  --deploy-config vllm_omni/deploy/moss_ttsd.yaml --port 8000\n```\n\n## Client usage\n\n```bash\ncurl -X POST http://localhost:8000/v1/audio/speech \\\n  -H \"Content-Type: application/json\" \\\n  -d '{\n    \"model\": \"OpenMOSS-Team/MOSS-TTSD-v1.0\",\n    \"input\": \"[S1] Hi, how was your day? [S2] Pretty good, thanks for asking!\",\n    \"voice\": \"default\",\n    \"response_format\": \"wav\"\n  }' --output dialogue.wav\n```\n\n## Known limitations\n\n- Output is **24 kHz mono**.\n- Shares the `OpenMOSS-Team/MOSS-Audio-Tokenizer` codec (~7 GB, auto-downloaded;\n  override with `MOSS_TTS_CODEC_PATH`).\n\n## References\n\n- [Model card](https://huggingface.co/OpenMOSS-Team/MOSS-TTSD-v1.0)\n- [MOSS-TTSD GitHub](https://github.com/OpenMOSS/MOSS-TTSD)\n- [MOSS-TTS Family](https://huggingface.co/OpenMOSS-Team/MOSS-TTS)\n",
  "source": {
    "source": "vllm-omni",
    "repo": "vllm-project/vllm-omni",
    "ref": "main",
    "sha": "cda4a1ef83b295ba332146a7709e6277725edc9f",
    "path": "recipes/models/OpenMOSS-Team/MOSS-TTSD-v1.0.yaml",
    "edit_url": "https://github.com/vllm-project/vllm-omni/edit/main/recipes/models/OpenMOSS-Team/MOSS-TTSD-v1.0.yaml",
    "overlays": []
  }
}