ACE Studio & StepFun: ACE-Step

ACE-Step is an open music foundation model co-led by ACE Studio and StepFun. Its diffusion-based architecture combines a compressed audio representation with a lightweight linear transformer, with the technical report emphasizing musical coherence and alignment between sung words and music. This version accepts style tags and optional lyrics as separate creative inputs, giving songwriters explicit control over the musical palette and words. Seed and duration controls also support instrumental compositions. Use cases: lyric-led song sketches, genre-tag exploration and music built around a defined song structure.

模型组织

ace-step

上下文长度

未提供

输入

文本

输出

音频

curl "$ROUTER_BASE_URL/api/v1/chat/completions" \
  -H "Authorization: Bearer $ROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "fal/fal-ai/ace-step",
  "messages": [
    {
      "role": "user",
      "content": "Gentle acoustic music with warm piano and soft guitar"
    }
  ],
  "modalities": [
    "text",
    "audio"
  ],
  "stream": true,
  "audio": {
    "format": "wav",
    "duration": 30
  }
}'