Initial commit: ComfyUI Text2Speech custom node with local TTS server

README.md

@@ -0,0 +1,32 @@
+# ComfyUI-Text2Speech
+
+A ComfyUI custom node for text-to-speech integration with a local TTS server.
+
+## Features
+
+- Custom ComfyUI node `HttpTTSToAudio` for sending text to a local TTS server.
+- Local TTS server using Coqui TTS (XTTS v2 model).
+- Easy setup with batch files for Windows.
+
+## Installation
+
+### TTS Server Setup
+
+1. Run `install.bat` to create a virtual environment and install dependencies.
+2. Run `run_tts.bat` to start the TTS server on `http://127.0.0.1:5002`.
+
+### ComfyUI Node
+
+1. Copy the `__init__.py` file to your ComfyUI `custom_nodes` folder (e.g., `ComfyUI/custom_nodes/ComfyUI_Text2Speech/__init__.py`).
+2. Install `requests` in your ComfyUI environment: `pip install requests`.
+3. Restart ComfyUI.
+
+## Usage
+
+- Use the `HttpTTSToAudio` node in ComfyUI under `Audio/TTS`.
+- Input text, language, and server URL.
+- Output is the path to the generated audio file.
+
+## Workflow
+
+Connect the audio path to audio/video processing nodes in ComfyUI for creating TTS-enhanced content.

__init__.py

@@ -0,0 +1,62 @@
+import os
+import json
+import requests
+import folder_paths
+
+class HttpTTSToAudio:
+    """
+    Sends text to a local TTS server and returns an audio file path.
+    Input: text (string), language (string)
+    Output: audio filepath (string)
+    """
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "text": ("STRING", {
+                    "multiline": True,
+                    "default": "Hello from ComfyUI."
+                }),
+                "language": ("STRING", {
+                    "default": "en"
+                }),
+                "server_url": ("STRING", {
+                    "default": "http://127.0.0.1:5002/tts"
+                })
+            }
+        }
+
+    RETURN_TYPES = ("STRING",)
+    RETURN_NAMES = ("audio_path",)
+    FUNCTION = "do_tts"
+    CATEGORY = "Audio/TTS"
+
+    def do_tts(self, text, language, server_url):
+        payload = {
+            "text": text,
+            "language": language
+        }
+        try:
+            resp = requests.post(server_url, json=payload, timeout=120)
+            resp.raise_for_status()
+            data = resp.json()
+            path = data.get("path")
+            if not path or not os.path.isfile(path):
+                raise RuntimeError(f"No file returned from TTS server: {data}")
+            # Copy to ComfyUI temp dir
+            with open(path, "rb") as f:
+                audio_bytes = f.read()
+            temp_dir = folder_paths.get_temp_directory()
+            out_name = os.path.basename(path)
+            out_path = os.path.join(temp_dir, out_name)
+            with open(out_path, "wb") as f:
+                f.write(audio_bytes)
+            return (out_path,)
+        except Exception as e:
+            print("TTS HTTP error:", e)
+            return ("",)
+
+NODE_CLASS_MAPPINGS = {
+    "HttpTTSToAudio": HttpTTSToAudio,
+}

install.bat

@@ -0,0 +1,16 @@
+@echo off
+echo Setting up TTS environment...
+
+if not exist venv (
+    python -m venv venv
+) else (
+    echo Virtual environment already exists.
+)
+
+call venv\Scripts\activate
+
+pip install --upgrade pip
+pip install -r requirements.txt
+
+echo TTS setup complete. Run run_tts.bat to start the server.
+pause

requirements.txt

@@ -0,0 +1,2 @@
+TTS
+flask

run_tts.bat

@@ -0,0 +1,12 @@
+@echo off
+echo Starting TTS server...
+
+if not exist venv (
+    echo Virtual environment not found. Run install.bat first.
+    pause
+    exit /b 1
+)
+
+call venv\Scripts\activate
+
+python tts_server.py

tts_server.py

@@ -0,0 +1,50 @@
+from flask import Flask, request, send_file, jsonify
+from TTS.api import TTS
+import os
+import uuid
+
+# Choose a model (example: XTTS v2 multilingual; you can pick any supported model)
+MODEL_NAME = "tts_models/multilingual/multi-dataset/xtts_v2"
+
+app = Flask(__name__)
+tts = TTS(MODEL_NAME)
+
+OUTPUT_DIR = "outputs"
+os.makedirs(OUTPUT_DIR, exist_ok=True)
+
+@app.route("/tts", methods=["POST"])
+def tts_route():
+    data = request.get_json()
+    text = data.get("text", "").strip()
+    speaker = data.get("speaker")  # optional
+    language = data.get("language", "en")  # optional
+
+    if not text:
+        return jsonify({"error": "No text"}), 400
+
+    out_name = f"{uuid.uuid4().hex}.wav"
+    out_path = os.path.join(OUTPUT_DIR, out_name)
+
+    # Generate audio
+    # For XTTS you can pass language, speaker_wav, etc.; adjust to your chosen model
+    tts.tts_to_file(
+        text=text,
+        file_path=out_path,
+        language=language
+    )
+
+    return jsonify({
+        "ok": True,
+        "file": out_name,
+        "path": out_path
+    })
+
+@app.route("/audio/<fname>", methods=["GET"])
+def get_audio(fname):
+    path = os.path.join(OUTPUT_DIR, fname)
+    if not os.path.isfile(path):
+        return jsonify({"error": "Not found"}), 404
+    return send_file(path, mimetype="audio/wav")
+
+if __name__ == "__main__":
+    app.run(host="127.0.0.1", port=5002)