音频转写 · Moss API

选择转写模式

场景	model	返回形式
普通音频转写	`moss-transcribe-1.0`	JSON / text
多说话人转写	`moss-transcribe-diarize-pro`	JSON，含 `segments`；可通过 `stream=true` 返回 SSE

请求参数

字段类型必填说明

model: string必填

模型 ID。普通转写使用 moss-transcribe-1.0；多说话人转写、热词和 SSE 流式输出使用 moss-transcribe-diarize-pro。

file: multipart file四选一

通过 multipart/form-data 直接上传本地音频或视频文件，无需先调用文件上传接口；不传 async=true 时同步返回转写结果。单文件最大 512 MB。

file_id: string四选一

引用 POST /v1/files 返回的文件 ID。

url: string四选一

服务端可访问的公网音频 URL；不得指向 localhost、回环地址或私网地址。

audio_url: string四选一

url 的兼容字段，访问限制相同；新调用推荐使用 url。

diarize: boolean可选

设为 true 时返回说话人分离结果，仅 moss-transcribe-diarize-pro 支持；不传或传 false 时只返回完整文本。

keyterms: string[]可选

热词列表，用于提升专有名词、人名、品牌名和业务术语的识别准确率。仅 moss-transcribe-diarize-pro 支持；最多 20 个，每个最多 30 个字符；不传或传 [] 表示不启用。使用 multipart/form-data 时以 JSON 字符串数组传入，例如 ["MOSS","SGLang"]。

stream: boolean可选

仅 moss-transcribe-diarize-pro 支持传 true，用于返回 SSE 转写事件；不支持同时传 async=true。

response_format: string可选

转写结果格式。非流式默认 json；流式固定返回 SSE 事件，建议显式传 verbose_json。

async: boolean可选

设为 true 时创建异步任务并立即返回任务对象；不能和 stream=true 同时使用。

webhook_url: string可选

异步任务完成后回调的 HTTPS URL。

支持格式与上传限制

项目	说明
单文件大小	最大 512 MB；直接上传与 `/v1/files` 保持一致。
已支持格式	AAC、AMR、FLAC、M4A、ALAC、MOV、MP3、MP4、MPG、OGG、OPUS、WAV、WebM、WMA
音视频容器	MOV、MP4、MPG、WebM均可完成转写

返回值

普通 JSON: object

response_format=json 且未启用 diarize 时返回完整转写文本。

说话人分离: object

启用 diarize，或使用 response_format=diarized_json 时，返回完整文本及说话人分段。

Text: string

response_format=text 时直接返回纯文本。

SSE: event stream

stream=true 时持续返回服务端事件，直到完成或失败；示例使用流式专用的 verbose_json。

异步任务: object

async=true 时立即返回任务对象，结果通过查询任务接口获取。

错误码

HTTPerror.code触发条件与处理建议

400: 4 个 error.code

输入来源、流式 / 异步组合或热词不合法。

错误说明

以下情况返回参数错误：

keyterms 不是字符串数组。
keyterms 包含非字符串元素。
普通 moss-transcribe-1.0 携带非空 keyterms。

下一步

上传文件

上传音频文件并获取 file_id。

查询音色列表

查询当前账号可用音色列表。

请求示例

curl -X POST "https://api.mosi.cn/v1/audio/transcriptions" \
  -H "Authorization: Bearer $MOSS_API_KEY" \
  -F "model=moss-transcribe-1.0" \
  -F "file=@./audio.mp3" \
  -F "response_format=json"

import os
import requests

with open("./audio.mp3", "rb") as file:
    response = requests.post(
        "https://api.mosi.cn/v1/audio/transcriptions",
        headers={
        "Authorization": f"Bearer {os.environ['MOSS_API_KEY']}",
        },
        data={
            "model": "moss-transcribe-1.0",
            "response_format": "json",
        },
        files={"file": file},
    )

response.raise_for_status()
print(response.json())

import { readFile } from "node:fs/promises"

const bytes = await readFile("./audio.mp3")
const form = new FormData()
form.append("model", "moss-transcribe-1.0")
form.append("file", new Blob([bytes]), "audio.mp3")
form.append("response_format", "json")

const response = await fetch("https://api.mosi.cn/v1/audio/transcriptions", {
  method: "POST",
  headers: {
    "Authorization": "Bearer " + process.env.MOSS_API_KEY,
  },
  body: form,
})

if (!response.ok) throw new Error(`HTTP ${response.status}`)
console.log(await response.json())

curl -X POST "https://api.mosi.cn/v1/audio/transcriptions" \
  -H "Authorization: Bearer $MOSS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "moss-transcribe-1.0",
  "file_id": "<file_id>",
  "response_format": "json"
}'

import os
import requests

response = requests.post(
        "https://api.mosi.cn/v1/audio/transcriptions",
        headers={
        "Authorization": f"Bearer {os.environ['MOSS_API_KEY']}",
        "Content-Type": "application/json",
        },
        json={
            "model": "moss-transcribe-1.0",
            "file_id": "<file_id>",
            "response_format": "json",
        },
)

response.raise_for_status()
print(response.json())

const response = await fetch("https://api.mosi.cn/v1/audio/transcriptions", {
  method: "POST",
  headers: {
    "Authorization": "Bearer " + process.env.MOSS_API_KEY,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "model": "moss-transcribe-1.0",
    "file_id": "<file_id>",
    "response_format": "json"
  }),
})

if (!response.ok) throw new Error(`HTTP ${response.status}`)
console.log(await response.json())

curl -X POST "https://api.mosi.cn/v1/audio/transcriptions" \
  -H "Authorization: Bearer $MOSS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "moss-transcribe-diarize-pro",
  "file_id": "<file_id>",
  "diarize": true,
  "response_format": "json"
}'

import os
import requests

response = requests.post(
        "https://api.mosi.cn/v1/audio/transcriptions",
        headers={
        "Authorization": f"Bearer {os.environ['MOSS_API_KEY']}",
        "Content-Type": "application/json",
        },
        json={
            "model": "moss-transcribe-diarize-pro",
            "file_id": "<file_id>",
            "diarize": True,
            "response_format": "json",
        },
)

response.raise_for_status()
print(response.json())

const response = await fetch("https://api.mosi.cn/v1/audio/transcriptions", {
  method: "POST",
  headers: {
    "Authorization": "Bearer " + process.env.MOSS_API_KEY,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "model": "moss-transcribe-diarize-pro",
    "file_id": "<file_id>",
    "diarize": true,
    "response_format": "json"
  }),
})

if (!response.ok) throw new Error(`HTTP ${response.status}`)
console.log(await response.json())

curl -X POST "https://api.mosi.cn/v1/audio/transcriptions" \
  -H "Authorization: Bearer $MOSS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "moss-transcribe-diarize-pro",
  "file_id": "<file_id>",
  "diarize": true,
  "response_format": "json",
  "keyterms": [
    "MOSS",
    "SGLang",
    "矩池云"
  ]
}'

import os
import requests

response = requests.post(
        "https://api.mosi.cn/v1/audio/transcriptions",
        headers={
        "Authorization": f"Bearer {os.environ['MOSS_API_KEY']}",
        "Content-Type": "application/json",
        },
        json={
            "model": "moss-transcribe-diarize-pro",
            "file_id": "<file_id>",
            "diarize": True,
            "response_format": "json",
            "keyterms": [
                "MOSS",
                "SGLang",
                "矩池云",
            ],
        },
)

response.raise_for_status()
print(response.json())

const response = await fetch("https://api.mosi.cn/v1/audio/transcriptions", {
  method: "POST",
  headers: {
    "Authorization": "Bearer " + process.env.MOSS_API_KEY,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "model": "moss-transcribe-diarize-pro",
    "file_id": "<file_id>",
    "diarize": true,
    "response_format": "json",
    "keyterms": [
      "MOSS",
      "SGLang",
      "矩池云"
    ]
  }),
})

if (!response.ok) throw new Error(`HTTP ${response.status}`)
console.log(await response.json())

curl -N -X POST "https://api.mosi.cn/v1/audio/transcriptions" \
  -H "Authorization: Bearer $MOSS_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Accept: text/event-stream" \
  -d '{
  "model": "moss-transcribe-diarize-pro",
  "file_id": "<file_id>",
  "diarize": true,
  "stream": true,
  "response_format": "verbose_json"
}'

import os
import requests

with requests.post(
        "https://api.mosi.cn/v1/audio/transcriptions",
        headers={
        "Authorization": f"Bearer {os.environ['MOSS_API_KEY']}",
        "Content-Type": "application/json",
        "Accept": "text/event-stream",
        },
        json={
            "model": "moss-transcribe-diarize-pro",
            "file_id": "<file_id>",
            "diarize": True,
            "stream": True,
            "response_format": "verbose_json",
        },
        stream=True,
    ) as response:
    response.raise_for_status()
    for line in response.iter_lines(decode_unicode=True):
        if line:
            print(line)

const response = await fetch("https://api.mosi.cn/v1/audio/transcriptions", {
  method: "POST",
  headers: {
    "Authorization": "Bearer " + process.env.MOSS_API_KEY,
    "Content-Type": "application/json",
    "Accept": "text/event-stream",
  },
  body: JSON.stringify({
    "model": "moss-transcribe-diarize-pro",
    "file_id": "<file_id>",
    "diarize": true,
    "stream": true,
    "response_format": "verbose_json"
  }),
})

if (!response.ok || !response.body) throw new Error(`HTTP ${response.status}`)
const reader = response.body.getReader()
const decoder = new TextDecoder()
while (true) {
  const { done, value } = await reader.read()
  if (done) break
  process.stdout.write(decoder.decode(value, { stream: true }))
}

响应示例

{
  "text": "This is a short transcription test audio."
}

{
  "task": "transcribe",
  "duration": 5.18,
  "text": "欢迎使用 Moss API。",
  "segments": [
    {
      "type": "transcript.text.segment",
      "id": "seg_001",
      "start": 0.16,
      "end": 5.18,
      "text": "欢迎使用 Moss API。",
      "speaker": "S01"
    }
  ]
}

data: {"type":"task.created","task_id":"task-123","object":"audio.transcription","status":"PROCESSING","model":"moss-transcribe-diarize-pro"}

data: {"type":"transcript.text.delta","delta":"欢迎"}

data: {"type":"transcript.segment.done","content_index":0,"speaker":"S01","start":0.16,"end":5.18,"text":"欢迎使用 Moss API。","session_usage":{"prompt_tokens":153,"completion_tokens":32,"total_tokens":185}}

data: {"type":"transcript.text.done","text":"欢迎使用 Moss API。","usage":{"prompt_tokens":153,"completion_tokens":32,"total_tokens":185}}

data: {"type":"task.created","task_id":"task-123","object":"audio.transcription","status":"PROCESSING","model":"moss-transcribe-diarize-pro"}

data: {"type":"transcript.text.delta","delta":"欢迎"}

data: {"type":"transcript.segment.done","content_index":0,"speaker":"S01","start":0.16,"end":5.18,"text":"欢迎使用 Moss API。","session_usage":{"prompt_tokens":153,"completion_tokens":32,"total_tokens":185}}

data: {"type":"transcript.text.done","text":"欢迎使用 Moss API。","usage":{"prompt_tokens":153,"completion_tokens":32,"total_tokens":185}}

data: {"type":"task.created","task_id":"task-123","object":"audio.transcription","status":"PROCESSING","model":"moss-transcribe-diarize-pro"}

data: {"type":"transcript.text.delta","delta":"欢迎"}

data: {"type":"transcript.segment.done","content_index":0,"speaker":"S01","start":0.16,"end":5.18,"text":"欢迎使用 Moss API。","session_usage":{"prompt_tokens":153,"completion_tokens":32,"total_tokens":185}}

data: {"type":"transcript.text.done","text":"欢迎使用 Moss API。","usage":{"prompt_tokens":153,"completion_tokens":32,"total_tokens":185}}