Skip to main content
Glama
kentaro

AivisSpeech MCP Server

by kentaro

aivis-speech-synthesis

Convert text to high-quality speech using customizable parameters like speaker ID, style, speed, pitch, and volume. Integrate the tool via the AivisSpeech MCP Server API for AI-powered voice synthesis.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
intonation_scaleNoイントネーションのスケール(1.0が標準)
output_sampling_rateNo出力音声のサンプリングレート(Hz)
pitch_scaleNo音高のスケール(1.0が標準)
post_phoneme_lengthNo音声の末尾の無音時間(秒)
pre_phoneme_lengthNo音声の先頭の無音時間(秒)
speaker_idNo音声合成に使用するスピーカーのID
speed_scaleNo話速のスケール(1.0が標準)
style_idNo音声合成に使用するスタイルのID
textYes音声合成するテキスト
volume_scaleNo音量のスケール(1.0が標準)

Implementation Reference

  • Registers the MCP tool 'aivis-speech-synthesis' using mcpServer.tool(), defining schema and handler inline.
    this.mcpServer.tool(
      MCP_MODEL_ID,
      {
        text: z.string().describe('音声合成するテキスト'),
        speaker_id: z.number().optional().describe('音声合成に使用するスピーカーのID'),
        style_id: z.number().optional().describe('音声合成に使用するスタイルのID'),
        speed_scale: z.number().min(0.5).max(2.0).optional().default(1.0).describe('話速のスケール(1.0が標準)'),
        pitch_scale: z.number().min(0.5).max(2.0).optional().default(1.0).describe('音高のスケール(1.0が標準)'),
        intonation_scale: z.number().min(0.0).max(2.0).optional().default(1.0).describe('イントネーションのスケール(1.0が標準)'),
        volume_scale: z.number().min(0.0).max(2.0).optional().default(1.0).describe('音量のスケール(1.0が標準)'),
        pre_phoneme_length: z.number().min(0.0).max(1.0).optional().default(0.1).describe('音声の先頭の無音時間(秒)'),
        post_phoneme_length: z.number().min(0.0).max(1.0).optional().default(0.1).describe('音声の末尾の無音時間(秒)'),
        output_sampling_rate: z.number().optional().default(24000).describe('出力音声のサンプリングレート(Hz)'),
      },
      async (params, extra) => {
        try {
          // デフォルトのスピーカーIDを.envから取得
          const defaultSpeakerId = parseInt(process.env.AIVIS_SPEECH_SPEAKER_ID || '888753760', 10);
    
          // AivisSpeech APIリクエストの作成
          const synthesisRequest: SynthesisRequest = {
            text: params.text,
            speaker: params.speaker_id || defaultSpeakerId,
            style_id: params.style_id,
            speed_scale: params.speed_scale,
            pitch_scale: params.pitch_scale,
            intonation_scale: params.intonation_scale,
            volume_scale: params.volume_scale,
            pre_phoneme_length: params.pre_phoneme_length,
            post_phoneme_length: params.post_phoneme_length,
            output_sampling_rate: params.output_sampling_rate,
          };
    
          try {
            // AivisSpeech APIを呼び出して音声合成を実行
            await aivisSpeechService.synthesize(synthesisRequest);
    
            // 正しいMCPレスポンス形式で返す
            return {
              content: [
                {
                  type: "text",
                  text: `「${params.text}」の音声合成が完了しました`
                }
              ]
            };
          } catch (synthesisError) {
            console.error('Synthesis error:', synthesisError);
            const errorMessage = synthesisError instanceof Error ? synthesisError.message : '音声合成処理中にエラーが発生しました';
            return {
              content: [{ type: "text", text: `音声合成に失敗しました: ${errorMessage}` }],
              isError: true
            };
          }
        } catch (error) {
          console.error('Request handling error:', error);
          const errorMessage = error instanceof Error ? error.message : '音声合成リクエストの処理中にエラーが発生しました';
          return {
            content: [{ type: "text", text: `音声合成に失敗しました: ${errorMessage}` }],
            isError: true
          };
        }
      }
    );
  • Zod input schema for the speech synthesis parameters.
    {
      text: z.string().describe('音声合成するテキスト'),
      speaker_id: z.number().optional().describe('音声合成に使用するスピーカーのID'),
      style_id: z.number().optional().describe('音声合成に使用するスタイルのID'),
      speed_scale: z.number().min(0.5).max(2.0).optional().default(1.0).describe('話速のスケール(1.0が標準)'),
      pitch_scale: z.number().min(0.5).max(2.0).optional().default(1.0).describe('音高のスケール(1.0が標準)'),
      intonation_scale: z.number().min(0.0).max(2.0).optional().default(1.0).describe('イントネーションのスケール(1.0が標準)'),
      volume_scale: z.number().min(0.0).max(2.0).optional().default(1.0).describe('音量のスケール(1.0が標準)'),
      pre_phoneme_length: z.number().min(0.0).max(1.0).optional().default(0.1).describe('音声の先頭の無音時間(秒)'),
      post_phoneme_length: z.number().min(0.0).max(1.0).optional().default(0.1).describe('音声の末尾の無音時間(秒)'),
      output_sampling_rate: z.number().optional().default(24000).describe('出力音声のサンプリングレート(Hz)'),
    },
  • Inline handler function that processes params, calls aivisSpeechService.synthesize(), and formats MCP response.
    async (params, extra) => {
      try {
        // デフォルトのスピーカーIDを.envから取得
        const defaultSpeakerId = parseInt(process.env.AIVIS_SPEECH_SPEAKER_ID || '888753760', 10);
    
        // AivisSpeech APIリクエストの作成
        const synthesisRequest: SynthesisRequest = {
          text: params.text,
          speaker: params.speaker_id || defaultSpeakerId,
          style_id: params.style_id,
          speed_scale: params.speed_scale,
          pitch_scale: params.pitch_scale,
          intonation_scale: params.intonation_scale,
          volume_scale: params.volume_scale,
          pre_phoneme_length: params.pre_phoneme_length,
          post_phoneme_length: params.post_phoneme_length,
          output_sampling_rate: params.output_sampling_rate,
        };
    
        try {
          // AivisSpeech APIを呼び出して音声合成を実行
          await aivisSpeechService.synthesize(synthesisRequest);
    
          // 正しいMCPレスポンス形式で返す
          return {
            content: [
              {
                type: "text",
                text: `「${params.text}」の音声合成が完了しました`
              }
            ]
          };
        } catch (synthesisError) {
          console.error('Synthesis error:', synthesisError);
          const errorMessage = synthesisError instanceof Error ? synthesisError.message : '音声合成処理中にエラーが発生しました';
          return {
            content: [{ type: "text", text: `音声合成に失敗しました: ${errorMessage}` }],
            isError: true
          };
        }
      } catch (error) {
        console.error('Request handling error:', error);
        const errorMessage = error instanceof Error ? error.message : '音声合成リクエストの処理中にエラーが発生しました';
        return {
          content: [{ type: "text", text: `音声合成に失敗しました: ${errorMessage}` }],
          isError: true
        };
      }
    }
  • synthesize() method implementing the core speech synthesis by calling AivisSpeech API endpoints /audio_query and /synthesis, generating WAV file, and playing it.
    async synthesize(params: SynthesisRequest): Promise<SynthesisResponse> {
      try {
        // 1. まずaudio_queryを取得
        const queryUrl = `${this.baseUrl}/audio_query`;
        const queryResponse = await axios.post(
          queryUrl,
          null,
          {
            params: {
              text: params.text,
              speaker: params.speaker
            }
          }
        );
    
        // 2. audio_queryを取得したら、必要に応じてパラメータを更新
        const audioQuery = queryResponse.data;
    
        if (params.style_id !== undefined) {
          audioQuery.style_id = params.style_id;
        }
    
        if (params.speed_scale !== undefined) {
          audioQuery.speed_scale = params.speed_scale;
        }
    
        if (params.pitch_scale !== undefined) {
          audioQuery.pitch_scale = params.pitch_scale;
        }
    
        if (params.intonation_scale !== undefined) {
          audioQuery.intonation_scale = params.intonation_scale;
        }
    
        if (params.volume_scale !== undefined) {
          audioQuery.volume_scale = params.volume_scale;
        }
    
        if (params.pre_phoneme_length !== undefined) {
          audioQuery.pre_phoneme_length = params.pre_phoneme_length;
        }
    
        if (params.post_phoneme_length !== undefined) {
          audioQuery.post_phoneme_length = params.post_phoneme_length;
        }
    
        if (params.output_sampling_rate !== undefined) {
          audioQuery.output_sampling_rate = params.output_sampling_rate;
        }
    
        // 3. 更新したaudio_queryを使って音声合成
        const synthesisUrl = `${this.baseUrl}/synthesis`;
        const synthesisResponse = await axios.post<ArrayBuffer>(
          synthesisUrl,
          audioQuery,
          {
            responseType: 'arraybuffer',
            params: {
              speaker: params.speaker
            },
            headers: {
              'Accept': 'audio/wav',
              'Content-Type': 'application/json'
            }
          }
        );
    
        // 音声データを一時ファイルに保存して再生
        const audioData = synthesisResponse.data;
        const tempDir = path.join(process.cwd(), 'temp');
    
        // 一時ディレクトリが存在しない場合は作成
        if (!fs.existsSync(tempDir)) {
          fs.mkdirSync(tempDir, { recursive: true });
        }
    
        // 一時ファイルのパスを生成
        const audioFilePath = path.join(tempDir, `speech_${Date.now()}.wav`);
    
        // 音声データをファイルに書き込み
        fs.writeFileSync(audioFilePath, Buffer.from(audioData));
    
        // node-wav-playerを使って音声を再生(メディアプレイヤーが立ち上がらない)
        try {
          await wavPlayer.play({
            path: audioFilePath,
            sync: false // 非同期再生
          });
        } catch (playError) {
          console.error('Error playing audio:', playError);
        }
    
        return {
          audioData: synthesisResponse.data
        };
      } catch (error) {
        console.error('Error in synthesize:', error);
        this.logDetailedError(error);
        throw new Error(`音声合成に失敗しました: ${this.getErrorMessage(error)}`);
      }
    }
  • SynthesisRequest interface defining the parameters passed to the synthesize helper.
    export interface SynthesisRequest {
      /**
       * 合成するテキスト
       */
      text: string;
    
      /**
       * 話者ID
       */
      speaker: number;
    
      /**
       * スタイルID
       */
      style_id?: number;
    
      /**
       * 話速のスケール(1.0が標準)
       */
      speed_scale?: number;
    
      /**
       * 音高のスケール(1.0が標準)
       */
      pitch_scale?: number;
    
      /**
       * イントネーションのスケール(1.0が標準)
       */
      intonation_scale?: number;
    
      /**
       * 音量のスケール(1.0が標準)
       */
      volume_scale?: number;
    
      /**
       * 音声の先頭の無音時間(秒)
       */
      pre_phoneme_length?: number;
    
      /**
       * 音声の末尾の無音時間(秒)
       */
      post_phoneme_length?: number;
    
      /**
       * 出力音声のサンプリングレート(Hz)
       */
      output_sampling_rate?: number;
    }
Behavior1/5

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

Tool has no description.

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

Conciseness1/5

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

Tool has no description.

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

Completeness1/5

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

Tool has no description.

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

Parameters1/5

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

Tool has no description.

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

Purpose1/5

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

Tool has no description.

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

Usage Guidelines1/5

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

Tool has no description.

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

Related 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/kentaro/aivis-speech-mcp'

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