{
  "family": "cohere_asr",
  "schema_version": 1,
  "display_name": "Cohere Transcribe",
  "description": "Cohere multilingual speech transcription with punctuation control.",
  "category": "asr",
  "status": "supported",
  "tasks": [
    "asr"
  ],
  "modes": [
    "offline"
  ],
  "languages": [
    "en",
    "fr",
    "de",
    "es",
    "it",
    "pt",
    "nl",
    "pl",
    "el",
    "ar",
    "ja",
    "zh",
    "vi",
    "ko"
  ],
  "capabilities": {},
  "dependencies": [],
  "options": {
    "request": [
      {
        "name": "language",
        "type": "string",
        "description": "Input audio language.",
        "required": false,
        "default": "en"
      },
      {
        "name": "punc",
        "type": "bool",
        "description": "Generate punctuation and capitalization.",
        "required": false,
        "default": true
      },
      {
        "name": "max_tokens",
        "type": "int",
        "description": "Maximum generated tokens per audio chunk.",
        "required": false,
        "min": 1,
        "max": 1014,
        "default": 256
      },
      {
        "name": "audio_chunk_mode",
        "type": "enum",
        "description": "Long-form audio chunking mode.",
        "values": [
          "auto",
          "quiet_energy",
          "fixed",
          "none"
        ],
        "required": false,
        "default": "auto"
      },
      {
        "name": "audio_chunk_duration_sec",
        "type": "float",
        "description": "Maximum duration of each audio chunk in seconds.",
        "required": false,
        "min": 0.04,
        "max": 35,
        "default": 35
      }
    ],
    "session": [
      {
        "name": "weight_type",
        "type": "enum",
        "description": "Matmul weight storage type.",
        "values": [
          "native",
          "f32",
          "f16",
          "bf16",
          "q8_0",
          "q4_0",
          "q4_1",
          "q5_0",
          "q5_1",
          "q2_k",
          "q3_k",
          "q4_k",
          "q5_k",
          "q6_k"
        ],
        "required": false,
        "default": "native"
      }
    ],
    "load": []
  },
  "runtime": {
    "tags": [
      "gguf",
      "cpu",
      "cuda"
    ]
  },
  "ui": {
    "tags": [
      "ASR",
      "GGUF"
    ],
    "docs": [
      "docs/models/cohere_asr.md"
    ],
    "recommended_package": "cohere_transcribe_bf16"
  },
  "package_defaults": {
    "download": {
      "kind": "huggingface_snapshot",
      "repo": "audio-cpp/audio.cpp-gguf",
      "revision": "main",
      "gated": false
    }
  },
  "packages": [
    {
      "id": "cohere_transcribe_bf16",
      "display_name": "Cohere Transcribe GGUF BF16",
      "default": true,
      "format": "gguf",
      "precision": "bf16",
      "target_directory": "Cohere-Transcribe-GGUF",
      "files": [
        "Cohere-Transcribe-GGUF/cohere-transcribe-03-2026-bf16.gguf"
      ],
      "strip_prefix": "Cohere-Transcribe-GGUF"
    },
    {
      "id": "cohere_transcribe_q8_0",
      "display_name": "Cohere Transcribe GGUF Q8_0",
      "format": "gguf",
      "precision": "q8_0",
      "target_directory": "Cohere-Transcribe-GGUF",
      "files": [
        "Cohere-Transcribe-GGUF/cohere-transcribe-03-2026-q8_0.gguf"
      ],
      "strip_prefix": "Cohere-Transcribe-GGUF"
    },
    {
      "id": "cohere_transcribe_q4_0",
      "display_name": "Cohere Transcribe GGUF Q4_0",
      "format": "gguf",
      "precision": "q4_0",
      "target_directory": "Cohere-Transcribe-GGUF",
      "files": [
        "Cohere-Transcribe-GGUF/cohere-transcribe-03-2026-q4_0.gguf"
      ],
      "strip_prefix": "Cohere-Transcribe-GGUF"
    }
  ],
  "sources": [
    {
      "format": "safetensors",
      "roots": {
        "model": "."
      },
      "files": {
        "config": "model:config.json",
        "tokenizer": "model:tokenizer.model"
      },
      "tensors": {
        "weights": "model:model.safetensors"
      }
    },
    {
      "format": "gguf",
      "roots": {
        "model": ".",
        "weights": "$gguf"
      },
      "files": {
        "config": "model:config.json",
        "tokenizer": "model:tokenizer.model"
      },
      "tensors": {
        "weights": "weights:"
      }
    }
  ]
}