SenseAudio Songmaker

Two-step song creation: generate lyrics → compose music. If the user already has lyrics, skip to Step 2.

Workflow

用户描述 → [Step 1] 生成歌词 → 用户确认/修改 → [Step 2] 生成歌曲 → 返回链接
                                ↑ 用户自带歌词时跳过 Step 1

Step 1: Generate Lyrics

Skip this step if the user provides their own lyrics.

LYRICS_RESP=$(curl -s -X POST "https://api.senseaudio.cn/v1/song/lyrics/create" \
  -H "Authorization: Bearer $SENSEAUDIO_API_KEY" \
  -H "Content-Type: application/json" \
  -d "{\"prompt\": \"\x3CUSER_PROMPT>\", \"provider\": \"sensesong\"}")

TASK_ID=$(echo $LYRICS_RESP | jq -r '.task_id // empty')

If task_id is present, poll until done:

while true; do
  POLL=$(curl -s "https://api.senseaudio.cn/v1/song/lyrics/pending/$TASK_ID" \
    -H "Authorization: Bearer $SENSEAUDIO_API_KEY")
  STATUS=$(echo $POLL | jq -r '.status')
  { [ "$STATUS" = "SUCCESS" ] || [ "$STATUS" = "FAILED" ]; } && break
  sleep 3
done
LYRICS=$(echo $POLL | jq -r '.response.data[0].text')

If no task_id (sync response), read directly:

LYRICS=$(echo $LYRICS_RESP | jq -r '.data[0].text')

Show the lyrics to the user and ask for confirmation or edits before proceeding.

Lyrics Format

The API uses structured section tags separated by ;:

[intro-medium] ; [verse] 第一段歌词内容 ; [chorus] 副歌内容 ; [bridge] 桥段 ; [outro-short]

Common tags: [intro-short] [intro-medium] [verse] [chorus] [bridge] [outro-short] [outro-medium] [inst-short]

If the user provides plain lyrics without tags, wrap them: [verse] \x3Clyrics> ; [chorus] \x3Cchorus>

Step 2: Generate Song

Build the request body from user preferences:

SONG_RESP=$(curl -s -X POST "https://api.senseaudio.cn/v1/song/music/create" \
  -H "Authorization: Bearer $SENSEAUDIO_API_KEY" \
  -H "Content-Type: application/json" \
  -d "{
    \"model\": \"sensesong\",
    \"lyrics\": \"\x3CLYRICS>\",
    \"title\": \"\x3CTITLE>\",
    \"vocal_gender\": \"\x3Cf|m>\",
    \"style\": \"\x3CSTYLE>\"
  }")

SONG_TASK=$(echo $SONG_RESP | jq -r '.task_id')

Song generation always returns a task_id. Poll until done (may take 30–120s):

while true; do
  POLL=$(curl -s "https://api.senseaudio.cn/v1/song/music/pending/$SONG_TASK" \
    -H "Authorization: Bearer $SENSEAUDIO_API_KEY")
  STATUS=$(echo $POLL | jq -r '.status')
  { [ "$STATUS" = "SUCCESS" ] || [ "$STATUS" = "FAILED" ]; } && break
  echo "生成中... ($STATUS)"
  sleep 5
done

Output

On SUCCESS, extract and display:

echo $POLL | jq -r '.response.data[0] | "标题：\(.title)\
时长：\(.duration) 秒\
音频：\(.audio_url)\
封面：\(.cover_url)"'

Example output:

标题：夏日物语
时长：187 秒
音频：https://cdn.senseaudio.cn/songs/xxx.mp3
封面：https://cdn.senseaudio.cn/covers/xxx.jpg

On FAILED, report the status and suggest retrying with a different style or simpler lyrics.