Saya sedang melakukan integrasi ke API gemini menggunakan Next.js dan Prisma. Ketika melakukan live audio, terjadi error seperti ini.

apa maksud dari 
 Error: [GoogleGenerativeAI Error]: Error fetching from https://generativelanguage.googleapis.com/v1beta/models/gemini-3-flash-preview:generateContent: [400 Bad Request] * GenerateContentRequest.generation_config.response_mime_type: allowed mimetypes are `text/plain`, `application/json`, `application/xml`, `application/yaml` and `text/x.enum`.

    at handleResponseNotOk (D:\overview-app\.next\dev\server\chunks\[root-of-the-server]__d680d9f5._.js:6560:11)
    at async makeRequest (D:\overview-app\.next\dev\server\chunks\[root-of-the-server]__d680d9f5._.js:6532:9)
    at async generateContent (D:\overview-app\.next\dev\server\chunks\[root-of-the-server]__d680d9f5._.js:6974:22)
    at async textToSpeech (lib\gemini.ts:73:18)
    at async POST (app\api\ask\route.ts:63:25)
  71 |   })
  72 |
> 73 |   const result = await model.generateContent({
     |                  ^
  74 |     contents: [{ role: "user", parts: [{text}] }],
  75 |     generationConfig: {
  76 |       responseMimeType: "audio/mpeg" {
  status: 400,
  statusText: 'Bad Request',
  errorDetails: undefined
}
 POST /api/ask 500 in 27.4s (compile: 6ms, proxy.ts: 15ms, render: 27.3s)