Skip to main content
Glama

generate_custom_song

Create songs by providing song title, lyrics, and style parameters to generate custom music with downloadable audio files.

Instructions

šŸŽµ Custom Mode: Generate songs based on detailed song information (user specifies song name, lyrics, style, etc.)

Use case: Use when users provide detailed song information including song name, complete lyrics, and style.

Example inputs:
- "Song name: Summer of Cicada Shedding, Lyrics: [complete lyrics], style: folk"

āš ļø COST WARNING: This tool makes an API call to MusicMCP.AI which may incur costs (5 credits per generation). Each generation creates 2 songs. Only use when explicitly requested by the user.

Language Note: Pass the title and lyrics in the user's input language.

Args:
    title (str): Song title, required
    lyric (str, optional): Complete lyrics content, not required when instrumental is True
    tags (str, optional): Music style tags (e.g., 'pop', 'rock', 'folk')
    instrumental (bool): Whether instrumental only (no lyrics)

Returns:
    Song information including download URLs

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
titleYes
instrumentalYes
lyricNo
tagsNo

Output Schema

TableJSON Schema
NameRequiredDescriptionDefault
resultYes

Implementation Reference

  • The main asynchronous handler function that implements the generate_custom_song tool. It handles input validation, API calls to MusicMCP.AI for custom song generation, polling for completion, and formatting the results as TextContent.
    async def generate_custom_song(
        title: str,
        instrumental: bool,
        lyric: str | None = None,
        tags: str | None = None
    ) -> list[TextContent]:
        try:
            if not api_key:
                raise Exception("Cannot find API key. Please set MUSICMCP_API_KEY environment variable.")
    
            if not title or title.strip() == "":
                raise Exception("Title is required.")
    
            # Lyric is only required when not instrumental
            if not instrumental and (not lyric or lyric.strip() == ""):
                raise Exception("Lyrics are required when instrumental is False.")
    
            url = f"{api_url}/music/generate/custom"
            headers = {
                'api-key': api_key,
                'Content-Type': 'application/json'
            }
    
            params = {
                "title": title,
                "instrumental": instrumental,
            }
    
            # Add optional parameters
            if lyric is not None and lyric.strip():
                params["lyric"] = lyric
    
            if tags is not None and tags.strip():
                params["tags"] = tags
    
            async with httpx.AsyncClient(timeout=httpx.Timeout(60.0)) as client:
                response = await client.post(url, json=params, headers=headers)
                response.raise_for_status()
                result = response.json()
    
            # API response format: {success, message, data}
            if not result or not result.get("success"):
                error_msg = result.get("message", "Unknown error") if result else "No response"
                raise Exception(f"Failed to create custom song generation task: {error_msg}")
    
            # Format: data contains {ids: [...]}
            data = result.get("data", {})
            song_ids = data.get("ids", [])
    
            if not song_ids:
                raise Exception("No song IDs returned from API")
    
            # Redirect debug info to stderr to avoid breaking JSON-RPC
            print(f"āœ… Custom song generation task created. Song IDs: {song_ids}", file=sys.stderr)
    
            # Poll for task completion
            current_timestamp = datetime.now().timestamp()
            while True:
                if (datetime.now().timestamp() - current_timestamp) > default_time_out:
                    raise Exception(f"Custom song generation timed out after {default_time_out} seconds")
    
                songs, status = await query_song_task(song_ids)
                # Redirect debug info to stderr to avoid breaking JSON-RPC
                print(f"šŸŽµ Custom song generation task status: {status}", file=sys.stderr)
                print(f"šŸŽµ Custom song generation task songs: {songs}", file=sys.stderr)
    
                if status == "error":
                    raise Exception("Custom song generation failed with error status")
                elif status == "timeout":
                    raise Exception("Custom song generation timed out")
                elif status == "completed" or status == "success":
                    break
                else:
                    time.sleep(2)
    
            # Return song information with URLs
            results = []
            for i, song in enumerate(songs, 1):
                song_url = song.get("songUrl") or song.get("audio_url") or song.get("url")
                song_title = song.get("songName", title).strip()
                song_id = song.get("id", "N/A")
    
                # Additional fields
                duration = song.get("duration", 0)
                tags = song.get("tags", "")
                img_url = song.get("imgUrl", "")
                lyric = song.get("lyric", "")
                instrumental_flag = song.get("instrumental", 0)
                created_at = song.get("createdAt", "")
    
                if not song_url:
                    continue
    
                # Format duration
                duration_str = f"{duration}s" if duration else "N/A"
    
                # Format instrumental status
                is_instrumental = "Yes" if instrumental_flag == 1 else "No"
    
                text = f"""āœ… Custom song '{title}' (version {i}) generated successfully!
    
    šŸ“Œ Title: {song_title}
    šŸ†” ID: {song_id}
    šŸ”— Download URL: {song_url}
    šŸ–¼ļø  Cover Image: {img_url if img_url else "N/A"}
    ā±ļø  Duration: {duration_str}
    šŸŽµ Style Tags: {tags if tags else "N/A"}
    šŸŽ¹ Instrumental: {is_instrumental}
    šŸ“… Created: {created_at if created_at else "N/A"}"""
    
                # Add lyrics if available and not instrumental
                if lyric and instrumental_flag == 0:
                    text += f"\nšŸ“ Lyrics:\n{lyric}"
    
                text += "\n\nYou can download or play the audio from the URL above."
    
                results.append(TextContent(type="text", text=text))
    
            if not results:
                raise Exception("No songs were generated successfully")
    
            return results
    
        except httpx.HTTPStatusError as e:
            error_detail = f"HTTP {e.response.status_code}"
            if e.response.status_code == 402:
                error_detail = "Insufficient credits. Please recharge your account."
            elif e.response.status_code == 401:
                error_detail = "Invalid API key. Please check your MUSICMCP_API_KEY."
            raise Exception(f"Request failed: {error_detail}") from e
        except Exception as e:
            raise Exception(f"Custom song generation failed: {str(e)}") from e
  • The @mcp.tool decorator registers the generate_custom_song tool with FastMCP, providing the schema via argument types and detailed usage description.
    @mcp.tool(
        description="""šŸŽµ Custom Mode: Generate songs based on detailed song information (user specifies song name, lyrics, style, etc.)
    
        Use case: Use when users provide detailed song information including song name, complete lyrics, and style.
    
        Example inputs:
        - "Song name: Summer of Cicada Shedding, Lyrics: [complete lyrics], style: folk"
    
        āš ļø COST WARNING: This tool makes an API call to MusicMCP.AI which may incur costs (5 credits per generation). Each generation creates 2 songs. Only use when explicitly requested by the user.
    
        Language Note: Pass the title and lyrics in the user's input language.
    
        Args:
            title (str): Song title, required
            lyric (str, optional): Complete lyrics content, not required when instrumental is True
            tags (str, optional): Music style tags (e.g., 'pop', 'rock', 'folk')
            instrumental (bool): Whether instrumental only (no lyrics)
    
        Returns:
            Song information including download URLs
        """
    )
  • Supporting helper function that queries the MusicMCP.AI API for the status of song generation tasks, used by generate_custom_song to poll until completion.
    async def query_song_task(song_ids: list[str]) -> tuple[list, str]:
        """Query song generation task status
    
        Args:
            song_ids: List of song IDs (batch query supported)
    
        Returns:
            Tuple of (songs_list, status_string)
        """
        try:
            url = f"{api_url}/music/generate/query"
            headers = {
                'api-key': api_key,
                'Content-Type': 'application/json'
            }
            params = {"ids": song_ids}
    
            async with httpx.AsyncClient(timeout=httpx.Timeout(60.0)) as client:
                response = await client.post(url, json=params, headers=headers)
                response.raise_for_status()
                result = response.json()
    
            # API response format: {success, message, data}
            if not result or not result.get("success"):
                return [], "error"
    
            # Format: data contains {songs: [...]}
            data = result.get("data", {})
            songs = data.get("songs", [])
    
            if songs and len(songs) > 0:
                all_complete = True
                any_error = False
    
                for song in songs:
                    status = song.get("status", 0)
                    if status == 0:  # 0 = Failed
                        any_error = True
                        break
                    elif status != 1:  # 1 = Completed, 2 = In Progress
                        all_complete = False
    
                if any_error:
                    return songs, "error"
                elif all_complete:
                    return songs, "completed"
                else:
                    return songs, "processing"
            else:
                return [], "processing"
    
        except Exception as e:
            # Redirect error info to stderr to avoid breaking JSON-RPC
            print(f"Query error: {str(e)}", file=sys.stderr)
            raise Exception(f"Failed to query song status: {str(e)}") from e
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries the full burden of behavioral disclosure. It effectively describes key traits: it's a generative tool that makes API calls to MusicMCP.AI, incurs costs (5 credits per generation), creates 2 songs per generation, and handles multilingual inputs. However, it doesn't mention rate limits, error conditions, or authentication needs.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is well-structured and appropriately sized. It starts with the core purpose, followed by usage guidelines, example, cost warning, language note, and parameter explanations. Every sentence adds value with no redundant information, making it easy to scan and understand.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's complexity (generative, cost-incurring, multilingual) and the presence of an output schema (which handles return values), the description is complete enough. It covers purpose, usage, behavioral traits, cost implications, language handling, and parameter semantics, providing comprehensive context for effective tool selection and invocation.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

With 0% schema description coverage, the description must compensate. It provides meaningful context for all parameters: title is required and part of detailed song information, lyric is optional and not required when instrumental is True, tags specify music style, and instrumental indicates whether the song is lyrics-free. The description adds value beyond the bare schema by explaining relationships between parameters.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool generates songs based on detailed song information, specifying the verb 'generate' and resource 'songs'. It distinguishes from sibling tools like generate_prompt_song by emphasizing 'detailed song information' and 'custom mode', making its specific purpose evident.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description explicitly states 'Use when users provide detailed song information including song name, complete lyrics, and style' and provides a clear example. It also includes a cost warning that advises 'Only use when explicitly requested by the user', offering strong guidance on when to use and when to avoid this tool.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/amCharlie/aimusic-mcp-tool'

If you have feedback or need assistance with the MCP directory API, please join our Discord server