Add transcribe

src/app/api/transcribe/app.py

@@ -0,0 +1,34 @@
+from fastapi import FastAPI, File, UploadFile
+from fastapi.middleware.cors import CORSMiddleware
+import whisper
+import os
+import tempfile
+
+app = FastAPI()
+model = whisper.load_model("base")  # Load the model once for efficiency
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["http://localhost:3000"],  # Frontend origin (adjust as needed)
+    allow_credentials=True,
+    allow_methods=["*"],  # Allow all HTTP methods (GET, POST, etc.)
+    allow_headers=["*"],  # Allow all headers (Authorization, Content-Type, etc.)
+)
+
+@app.post("/transcribe")
+async def transcribe_audio(file: UploadFile = File(...)):
+    # Save the uploaded file to a temporary location
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as temp_file:
+        temp_file.write(await file.read())
+        temp_path = temp_file.name
+
+    try:
+        # Transcribe the audio
+        result = model.transcribe("inputs/test.mp3")
+        transcription = result["text"]
+        print(transcription)
+    finally:
+        # Clean up temporary file
+        os.remove(temp_path)
+
+    return {"transcription": transcription}

src/app/api/transcribe/route.ts

@@ -0,0 +1,55 @@
+import { NextRequest, NextResponse } from "next/server";
+import multer from "multer";
+import fs from "fs/promises";
+import whisper from "openai-whisper";
+
+const upload = multer({ dest: "uploads/" });
+
+export const config = {
+  api: {
+    bodyParser: false, // Disable Next.js's body parsing for file uploads
+  },
+};
+
+// Whisper model (initialize once for efficiency)
+const model = whisper.load_model("base");
+
+// Utility to transcribe audio
+async function transcribeAudio(filePath: string): Promise<string> {
+  const transcription = await model.transcribe(filePath);
+  return transcription.text;
+}
+
+async function parseMultipartForm(req: NextRequest): Promise<File> {
+  return new Promise((resolve, reject) => {
+    const multerMiddleware = upload.single("audio");
+    multerMiddleware(req as any, {} as any, (error: any) => {
+      if (error) return reject(error);
+      resolve(req.file);
+    });
+  });
+}
+
+export async function POST(req: NextRequest) {
+  try {
+    // Parse the incoming multipart form data
+    const file = await parseMultipartForm(req);
+
+    if (!file) {
+      return NextResponse.json({ error: "No audio file provided" }, { status: 400 });
+    }
+
+    const filePath = file.path;
+
+    // Transcribe the audio
+    const transcription = await transcribeAudio(filePath);
+
+    // Clean up the uploaded file
+    await fs.unlink(filePath);
+
+    return NextResponse.json({ transcription });
+  } catch (error) {
+    console.error("Error during transcription:", error);
+    return NextResponse.json({ error: "Transcription failed" }, { status: 500 });
+  }
+}