在 Hermes Agent 中使用 Suno。
Suno 从文本提示词或精确歌词生成完整歌曲——人声、乐器和混音。v4 到 v5.5 各版本均可通过 RunAPI 使用，由于 Suno 没有官方公开 API，RunAPI 提供了 API 访问能力。Hermes Agent 通过 custom:runapi 提供商，使用与聊天相同的 API 密钥调用它。
使用 RunAPI 通过 Suno 生成歌曲。
要求：
- 从 RUNAPI_API_KEY 读取 API 密钥。
- 使用 custom:runapi 提供商，base_url 为 https://runapi.ai/v1。
- 调用 POST https://runapi.ai/api/v1/suno/text_to_music
- 将 model 设为 "suno-v5.5"。
- 将 vocal_mode 设为 "auto_lyrics"，以根据提示词由 AI 生成歌词。
- 将 prompt 设为描述你想要的歌曲。
- 该任务为异步任务。轮询返回的 task_id 直到 status 为 "completed"。
- 完成后，从响应 output 中读取音频 URL。
curl -X POST https://runapi.ai/api/v1/suno/text_to_music \
-H "Authorization: Bearer $RUNAPI_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "suno-v5.5",
"vocal_mode": "auto_lyrics",
"prompt": "An upbeat indie pop track about coding late at night, warm synths and acoustic guitar",
"vocal_gender": "female"
}'
{
"task_id": "tsk_abc123",
"status": "pending",
"model": "suno-v5.5"
}
三步在 Hermes Agent 中使用 Suno
Configure RunAPI
Set RUNAPI_API_KEY in the environment where Hermes Agent runs. If you already added RunAPI as a custom:runapi provider, the same key and base_url handle Suno music requests.
export RUNAPI_API_KEY=runapi_xxx
Call text_to_music
Send a POST to the text_to_music endpoint with the model and vocal_mode. Use auto_lyrics with a prompt for AI-written lyrics, exact_lyrics with lyrics/style/title for your own words, or instrumental with style/title for no vocals.
POST /api/v1/suno/text_to_music
Poll for the result
The endpoint returns a task_id immediately. Song generation takes 30 seconds to several minutes depending on version. Poll the task status endpoint until the status is completed, then read the output audio URL.
GET /api/v1/suno/text_to_music/tsk_abc123
Suno text_to_music API 参数
| 参数 | 类型 | 说明 |
|---|---|---|
model |
string |
Required. suno-v4, suno-v4.5, suno-v4.5-all, suno-v4.5-plus, suno-v5, or suno-v5.5. |
vocal_mode |
string |
Required. auto_lyrics (AI writes lyrics from prompt), exact_lyrics (you provide lyrics/style/title), or instrumental (no vocals, requires style/title). |
prompt |
string |
Song brief for auto_lyrics mode. Describe genre, mood, tempo, and subject. Forbidden in exact_lyrics and instrumental modes. |
lyrics |
string |
Exact lyrics to sing. Required for exact_lyrics mode. Forbidden in auto_lyrics and instrumental modes. |
style |
string |
Music style description, e.g. lo-fi hip hop, 80s synthwave. Required for exact_lyrics and instrumental modes. |
title |
string |
Song title. Required for exact_lyrics and instrumental modes. |
vocal_gender |
string |
Optional. male or female. |
negative_tags |
string |
Optional. Styles to avoid, e.g. heavy metal, screaming. |
duration_seconds |
integer |
Optional. Target duration in seconds. |
persona_id |
string |
Optional. Persona ID for custom voice or style profiles. |
persona_type |
string |
Optional. style or voice. Selects the persona category. |
callback_url |
string |
Optional. Webhook URL that receives a POST when the task completes. |
Hermes Agent 上的 Suno 是什么？
Suno 是一个文生音乐生成器，能从文字描述或歌词生成包含人声、乐器和混音的完整歌曲。Hermes Agent 通过 custom:runapi provider 调用它，适合多步骤内容工作流——先用 Kling 生成视频素材，再用 Suno 创作匹配的原创配乐，在单次 agent 运行中完成。
Suno 使用场景
正式录制前的作曲草稿
描述一个歌曲概念——曲风、情绪、节奏、主题——即可获得完整小样。快速迭代创意，在进录音棚前听到不同风格和编曲的效果。
AI 翻唱不同风格
通过 cover_audio 端点对现有曲目重新编曲。改变曲风、替换乐器或调整情绪，同时保留核心旋律和结构。
完整视听内容流水线
在 Hermes Agent 工作流中将 Suno 与视频模型串联——生成视频素材，再创作匹配的原创配乐，产出无需单独音乐授权即可直接发布的内容。
Suno + Hermes Agent 常见问题
Be specific about genre, mood, tempo, and instrumentation in your prompt. For example, "upbeat 120 BPM indie pop with acoustic guitar and warm synths" works better than "happy song." Use negative_tags to exclude unwanted elements. Meta tags in the prompt help Suno understand song structure.
Add a custom:runapi provider in the Hermes Agent config with base_url set to https://runapi.ai/v1 and api_key reading from RUNAPI_API_KEY. The same provider handles chat, image, video, and music endpoints -- no separate config for Suno.
Yes. Use the extend_music endpoint to continue a song from a specific timestamp. You can chain multiple extensions to build longer tracks. RunAPI also offers cover_audio to restyle existing tracks and create_mashup to combine elements from multiple songs.
Yes. Set vocal_mode to instrumental and provide style and title fields. The prompt and lyrics fields must be omitted in instrumental mode. The API enforces these cross-field rules and returns a validation error if conflicting fields are sent.
RunAPI exposes cover_audio (restyle an existing track), extend_music (continue a song from a timestamp), create_mashup (combine elements from multiple tracks), text_to_sound (sound effects), and utility endpoints like generate_lyrics, generate_persona, and separate_audio_stems. All share the same API key and task lifecycle.
Set the is_custom parameter to true and provide your exact lyrics in the prompt field when calling the Suno endpoint through Hermes Agent. The custom:runapi provider passes all parameters to Suno, which generates the song using your lyrics.
Yes. Hermes Agent can generate video with Kling or Veo 3, then create a matching soundtrack with Suno, combining both outputs into a complete audiovisual package within one workflow.