speechflow 1.7.0 → 2.0.0

package/README.md

@@ -26,7 +26,8 @@ speech-to-speech).
 **SpeechFlow** comes with built-in graph nodes for various functionalities:
 
 - file and audio device I/O for local connectivity,
-- WebSocket and MQTT network I/O for remote connectivity,
+- WebSocket, MQTT, VBAN, and WebRTC network I/O for remote connectivity,
+- external command execution I/O for process integration,
 - local Voice Activity Detection (VAD),
 - local voice gender recognition,
 - local audio LUFS-S/RMS metering,
@@ -38,20 +39,27 @@ speech-to-speech).
 - remote-controlable audio muting,
 - cloud-based speech-to-text conversion with
   [Amazon Transcribe](https://aws.amazon.com/transcribe/),
-  [OpenAI GPT-Transcribe](https://platform.openai.com/docs/models/gpt-4o-mini-transcribe), or
-  [Deepgram](https://deepgram.com).
+  [OpenAI GPT-Transcribe](https://platform.openai.com/docs/models/gpt-4o-mini-transcribe),
+  [Deepgram](https://deepgram.com), or
+  [Google Cloud Speech-to-Text](https://cloud.google.com/speech-to-text).
 - cloud-based text-to-text translation (or spelling correction) with
   [DeepL](https://deepl.com),
   [Amazon Translate](https://aws.amazon.com/translate/),
-  [Google Cloud Translate](https://cloud.google.com/translate), or
-  [OpenAI GPT](https://openai.com).
+  [Google Cloud Translate](https://cloud.google.com/translate),
+  [OpenAI GPT](https://openai.com),
+  [Anthropic Claude](https://anthropic.com), or
+  [Google Gemini](https://ai.google.dev).
 - local text-to-text translation (or spelling correction) with
-  [Ollama/Gemma](https://ollama.com) or
-  [Transformers/OPUS](https://github.com/Helsinki-NLP/Opus-MT).
+  [Ollama](https://ollama.com) or
+  [OPUS-MT](https://github.com/Helsinki-NLP/Opus-MT).
 - cloud-based text-to-speech conversion with
-  [ElevenLabs](https://elevenlabs.io/) or
-  [Amazon Polly](https://aws.amazon.com/polly/).
-- local text-to-speech conversion with [Kokoro](https://github.com/nazdridoy/kokoro-tts).
+  [OpenAI TTS](https://platform.openai.com/docs/guides/text-to-speech),
+  [ElevenLabs](https://elevenlabs.io/),
+  [Amazon Polly](https://aws.amazon.com/polly/), or
+  [Google Cloud Text-to-Speech](https://cloud.google.com/text-to-speech).
+- local text-to-speech conversion with
+  [Kokoro](https://github.com/nazdridoy/kokoro-tts) or
+  [Supertonic](https://huggingface.co/Supertone/supertonic).
 - local [FFmpeg](https://ffmpeg.org/)-based speech-to-speech conversion,
 - local WAV speech-to-speech decoding/encoding,
 - local text-to-text formatting, regex-based modification,
@@ -221,8 +229,8 @@ They can also be found in the sample [speechflow.yaml](./etc/speechflow.yaml) fi
 
   ```
   xio-device(device: env.SPEECHFLOW_DEVICE_MIC, mode: "r") |
-      a2a-wav(mode: "encode") |
-          xio-file(path: "capture.wav", mode: "w", type: "audio")
+      a2a-wav(mode: "encode", seekable: true) |
+          xio-file(path: "capture.wav", mode: "w", type: "audio", seekable: true)
   ```
 
 - **Pass-Through**: Pass-through audio from microphone device to speaker
@@ -335,7 +343,10 @@ First a short overview of the available processing nodes:
   **xio-file**,
   **xio-device**,
   **xio-websocket**,
-  **xio-mqtt**.
+  **xio-mqtt**,
+  **xio-vban**,
+  **xio-webrtc**,
+  **xio-exec**.
 - Audio-to-Audio nodes:
   **a2a-ffmpeg**,
   **a2a-wav**,
@@ -353,22 +364,29 @@ First a short overview of the available processing nodes:
 - Audio-to-Text nodes:
   **a2t-openai**,
   **a2t-amazon**,
-  **a2t-deepgram**.
+  **a2t-deepgram**,
+  **a2t-google**.
 - Text-to-Text nodes:
   **t2t-deepl**,
   **t2t-amazon**,
-  **t2t-openai**,
-  **t2t-ollama**,
-  **t2t-transformers**,
+  **t2t-opus**,
   **t2t-google**,
+  **t2t-translate**,
+  **t2t-spellcheck**,
+  **t2t-punctuation**,
   **t2t-modify**,
+  **t2t-profanity**,
+  **t2t-summary**,
   **t2t-subtitle**,
   **t2t-format**,
   **t2t-sentence**.
 - Text-to-Audio nodes:
+  **t2a-openai**,
   **t2a-amazon**,
   **t2a-elevenlabs**,
-  **t2a-kokoro**.
+  **t2a-google**,
+  **t2a-kokoro**,
+  **t2a-supertonic**.
 - Any-to-Any nodes:
   **x2x-filter**,
   **x2x-trace**.
@@ -384,20 +402,24 @@ external files, devices and network services.
 
   > This node allows the reading/writing from/to files or from StdIO. It
   > is intended to be used as source and sink nodes in batch processing,
-  > and as sing nodes in real-time processing.
+  > and as sing nodes in real-time processing. When `seekable` is enabled
+  > for write mode, the node uses a file descriptor allowing random access
+  > writes to specific file positions via the `chunk:seek` metadata field.
+  > Option `seekable` cannot be used on StdIO.
 
   | Port    | Payload     |
   | ------- | ----------- |
   | input   | text, audio |
   | output  | text, audio |
 
-  | Parameter  | Position  | Default  | Requirement           |
-  | ---------- | --------- | -------- | --------------------- |
-  | **path**   | 0         | *none*   | *none*                |
-  | **mode**   | 1         | "r"      | `/^(?:r\|w\|rw)$/`    |
-  | **type**   | 2         | "audio"  | `/^(?:audio\|text)$/` |
-  | **chunka** |           | 200      | `10 <= n <= 1000`     |
-  | **chunkt** |           | 65536    | `1024 <= n <= 131072` |
+  | Parameter      | Position  | Default  | Requirement           |
+  | -------------- | --------- | -------- | --------------------- |
+  | **path**       | 0         | *none*   | *none*                |
+  | **mode**       | 1         | "r"      | `/^(?:r\|w)$/`        |
+  | **type**       | 2         | "audio"  | `/^(?:audio\|text)$/` |
+  | **seekable**   |           | false    | *none*                |
+  | **chunkAudio** |           | 200      | `10 <= n <= 1000`     |
+  | **chunkText**  |           | 65536    | `1024 <= n <= 131072` |
 
 - Node: **xio-device**<br/>
   Purpose: **Microphone/speaker device source/sink**<br/>
@@ -437,11 +459,12 @@ external files, devices and network services.
   | ----------- | --------- | -------- | --------------------- |
   | **listen**  | *none*    | *none*   | `/^(?:\|ws:\/\/(.+?):(\d+))$/` |
   | **connect** | *none*    | *none*   | `/^(?:\|ws:\/\/(.+?):(\d+)(?:\/.*)?)$/` |
-  | **type**    | *none*    | "audio"  | `/^(?:audio\|text)$/` |
+  | **mode**    | *none*    | "r"      | `/^(?:r\|w\|rw)$/`    |
+  | **type**    | *none*    | "text"   | `/^(?:audio\|text)$/` |
 
 - Node: **xio-mqtt**<br/>
-  Purpose: **MQTT sink**<br/>
-  Example: `xio-mqtt(url: "mqtt://127.0.0.1:1883", username: "foo", password: "bar", topic: "quux")`
+  Purpose: **MQTT source/sink**<br/>
+  Example: `xio-mqtt(url: "mqtt://127.0.0.1:1883", username: "foo", password: "bar", topicWrite: "quux")`
   Notice: this node requires a peer MQTT broker!
 
   > This node allows reading/writing from/to MQTT broker topics. It is
@@ -450,15 +473,94 @@ external files, devices and network services.
 
   | Port    | Payload     |
   | ------- | ----------- |
-  | input   | text        |
-  | output  | none        |
+  | input   | text, audio |
+  | output  | text, audio |
 
-  | Parameter    | Position  | Default  | Requirement           |
-  | ------------ | --------- | -------- | --------------------- |
-  | **url**      | 0         | *none*   | `/^(?:\|(?:ws\|mqtt):\/\/(.+?):(\d+))$/` |
-  | **username** | 1         | *none*   | `/^.+$/` |
-  | **password** | 2         | *none*   | `/^.+$/` |
-  | **topic**    | 3         | *none*   | `/^.+$/` |
+  | Parameter      | Position  | Default  | Requirement           |
+  | -------------- | --------- | -------- | --------------------- |
+  | **url**        | 0         | *none*   | `/^(?:\|(?:ws\|mqtt):\/\/(.+?):(\d+)(?:\/.*)?)$/` |
+  | **username**   | 1         | *none*   | `/^.+$/` |
+  | **password**   | 2         | *none*   | `/^.+$/` |
+  | **topicRead**  | 3         | *none*   | `/^.+$/` |
+  | **topicWrite** | 4         | *none*   | `/^.+$/` |
+  | **mode**       | 5         | "w"      | `/^(?:r\|w\|rw)$/` |
+  | **type**       | 6         | "text"   | `/^(?:audio\|text)$/` |
+
+- Node: **xio-vban**<br/>
+  Purpose: **VBAN network audio source/sink**<br/>
+  Example: `xio-vban(listen: 6980, stream: "Stream1", mode: "r")`
+  Notice: this node requires a peer VBAN-compatible application!
+
+  > This node allows reading/writing audio from/to VBAN (VoiceMeeter
+  > Audio Network) protocol endpoints. It is intended to be used for
+  > real-time audio streaming with applications like VoiceMeeter,
+  > VB-Audio Matrix, or other VBAN-compatible software. It supports
+  > various audio bit resolutions (8-bit, 16-bit, 24-bit, 32-bit,
+  > float32, float64) and automatic channel downmixing to mono.
+
+  | Port    | Payload     |
+  | ------- | ----------- |
+  | input   | audio       |
+  | output  | audio       |
+
+  | Parameter   | Position  | Default   | Requirement                  |
+  | ----------- | --------- | --------- | ---------------------------- |
+  | **listen**  | 0         | ""        | `/^(?:\|\d+\|.+?:\d+)$/`     |
+  | **connect** | 1         | ""        | `/^(?:\|.+?:\d+)$/`          |
+  | **stream**  | 2         | "Stream"  | `/^.{1,16}$/`                |
+  | **mode**    | 3         | "rw"      | `/^(?:r\|w\|rw)$/`           |
+
+- Node: **xio-webrtc**<br/>
+  Purpose: **WebRTC audio streaming source (WHIP) or sink (WHEP)**<br/>
+  Example: `xio-webrtc(listen: 8085, path: "/webrtc", mode: "r")`
+
+  > This node allows real-time audio streaming using WebRTC technology
+  > via WebRTC-HTTP Ingestion Protocol (WHIP) or WebRTC-HTTP Egress
+  > Protocol (WHEP). It provides an HTTP server for SDP negotiation
+  > and uses Opus codec for audio encoding/decoding at 48kHz. The node
+  > can operate in WHIP mode (i.e., read mode where publishers POST
+  > SDP offers to SpeechFlow and SpeechFlow receives audio stream from
+  > them) or WHEP mode (i.e., write mode where viewers POST SDP offers
+  > to SpeechFlow and SpeechFlow sends audio stream to them). This node
+  > supports multiple simultaneous connections, configurable ICE servers
+  > for NAT traversal, and automatic connection lifecycle management.
+
+  | Port    | Payload     |
+  | ------- | ----------- |
+  | input   | audio       |
+  | output  | audio       |
+
+  | Parameter      | Position  | Default   | Requirement                  |
+  | -------------- | --------- | --------- | ---------------------------- |
+  | **listen**     | 0         | "8085"    | `/^(?:\d+\|.+?:\d+)$/`       |
+  | **path**       | 1         | "/webrtc" | `/^\/.+$/`                   |
+  | **mode**       | 2         | "r"       | `/^(?:r\|w)$/`               |
+  | **iceServers** | 3         | ""        | `/^.*$/`                     |
+
+- Node: **xio-exec**<br/>
+  Purpose: **External command execution source/sink**<br/>
+  Example: `xio-exec(command: "ffmpeg -i - -f s16le -", mode: "rw", type: "audio")`
+
+  > This node allows reading/writing from/to external commands via stdin/stdout.
+  > It executes arbitrary commands and pipes audio or text data through them,
+  > enabling integration with external processing tools. The node supports
+  > read-only mode (capturing stdout), write-only mode (sending to stdin),
+  > and bidirectional mode (both stdin and stdout). This is useful for integrating
+  > external audio/text processing tools like FFmpeg, SoX, or custom scripts into
+  > the SpeechFlow pipeline.
+
+  | Port    | Payload     |
+  | ------- | ----------- |
+  | input   | text, audio |
+  | output  | text, audio |
+
+  | Parameter      | Position  | Default  | Requirement           |
+  | -------------- | --------- | -------- | --------------------- |
+  | **command**    | 0         | *none*   | *required*            |
+  | **mode**       | 1         | "r"      | `/^(?:r\|w\|rw)$/`    |
+  | **type**       | 2         | "audio"  | `/^(?:audio\|text)$/` |
+  | **chunkAudio** |           | 200      | `10 <= n <= 1000`     |
+  | **chunkText**  |           | 65536    | `1024 <= n <= 131072` |
 
 ### Audio-to-Audio Nodes
 
@@ -477,10 +579,10 @@ The following nodes process audio chunks only.
   | input   | audio       |
   | output  | audio       |
 
-  | Parameter   | Position  | Default  | Requirement        |
-  | ----------- | --------- | -------- | ------------------ |
-  | **src**     | 0         | "pcm"    | `/^(?:pcm\|wav\|mp3\|opus)$/` |
-  | **dst**     | 1         | "wav"    | `/^(?:pcm\|wav\|mp3\|opus)$/` |
+  | Parameter | Position  | Default  | Requirement        |
+  | --------- | --------- | -------- | ------------------ |
+  | **src**   | 0         | "pcm"    | `/^(?:pcm\|wav\|mp3\|opus)$/` |
+  | **dst**   | 1         | "wav"    | `/^(?:pcm\|wav\|mp3\|opus)$/` |
 
 - Node: **a2a-wav**<br/>
   Purpose: **WAV audio format conversion**<br/>
@@ -489,15 +591,20 @@ The following nodes process audio chunks only.
   > This node allows converting between PCM and WAV audio formats. It is
   > primarily intended to support the reading/writing of external WAV
   > format files, although SpeechFlow internally uses PCM format only.
+  > When `seekable` is enabled in encode mode, the node writes a corrected
+  > WAV header at the end of processing with accurate file size information
+  > by seeking back to position 0, producing standard-compliant WAV files.
+  > Option `seekable` requires a seekable output stream.
 
   | Port    | Payload     |
   | ------- | ----------- |
   | input   | audio       |
   | output  | audio       |
 
-  | Parameter   | Position  | Default  | Requirement              |
-  | ----------- | --------- | -------- | ------------------------ |
-  | **mode**    | 0         | "encode" | `/^(?:encode\|decode)$/` |
+  | Parameter    | Position  | Default  | Requirement              |
+  | ------------ | --------- | -------- | ------------------------ |
+  | **mode**     | 0         | "encode" | `/^(?:encode\|decode)$/` |
+  | **seekable** | 1         | false    | *none*                   |
 
 - Node: **a2a-mute**<br/>
   Purpose: **volume muting node**<br/>
@@ -512,8 +619,8 @@ The following nodes process audio chunks only.
   | input   | audio       |
   | output  | audio       |
 
-  | Parameter   | Position  | Default  | Requirement              |
-  | ----------- | --------- | -------- | ------------------------ |
+  | Parameter | Position  | Default  | Requirement              |
+  | --------- | --------- | -------- | ------------------------ |
 
 - Node: **a2a-meter**<br/>
   Purpose: **Loudness metering node**<br/>
@@ -531,7 +638,7 @@ The following nodes process audio chunks only.
 
   | Parameter     | Position  | Default  | Requirement            |
   | ------------- | --------- | -------- | ---------------------- |
-  | **interval**  | 0         | 250      | *none*                 |
+  | **interval**  | 0         | 100      | *none*                 |
   | **mode**      | 1         | "filter" | `/^(?:filter\|sink)$/` |
   | **dashboard** |           | *none*   | *none*                 |
 
@@ -548,8 +655,8 @@ The following nodes process audio chunks only.
   | input   | audio       |
   | output  | audio       |
 
-  | Parameter   | Position  | Default  | Requirement              |
-  | ----------- | --------- | -------- | ------------------------ |
+  | Parameter | Position  | Default  | Requirement              |
+  | --------- | --------- | -------- | ------------------------ |
   | **mode**               | *none* | "unplugged" | `/^(?:silenced\|unplugged)$/` |
   | **posSpeechThreshold** | *none* | 0.50  | *none* |
   | **negSpeechThreshold** | *none* | 0.35  | *none* |
@@ -571,11 +678,12 @@ The following nodes process audio chunks only.
   | input   | audio       |
   | output  | audio       |
 
-  | Parameter      | Position  | Default  | Requirement              |
-  | -------------- | --------- | -------- | ------------------------ |
-  | **window**     | 0         | 500      | *none*                   |
-  | **treshold**   | 1         | 0.50     | *none*                   |
-  | **hysteresis** | 2         | 0.25     | *none*                   |
+  | Parameter           | Position  | Default  | Requirement              |
+  | ------------------- | --------- | -------- | ------------------------ |
+  | **window**          | 0         | 500      | *none*                   |
+  | **threshold**       | 1         | 0.50     | *none*                   |
+  | **hysteresis**      | 2         | 0.25     | *none*                   |
+  | **volumeThreshold** | 3         | -45      | *none*                   |
 
 - Node: **a2a-speex**<br/>
   Purpose: **Speex Noise Suppression node**<br/>
@@ -590,9 +698,9 @@ The following nodes process audio chunks only.
   | input   | audio       |
   | output  | audio       |
 
-  | Parameter   | Position  | Default  | Requirement              |
-  | ----------- | --------- | -------- | ------------------------ |
-  | **attentuate** | 0 | -18  | *none* | `-60 <= n <= 0` |
+  | Parameter      | Position  | Default  | Requirement        |
+  | -------------- | --------- | -------- | ------------------ |
+  | **attentuate** | 0         | -18      | `-60 <= n <= 0`    |
 
 - Node: **a2a-rnnoise**<br/>
   Purpose: **RNNoise Noise Suppression node**<br/>
@@ -606,8 +714,8 @@ The following nodes process audio chunks only.
   | input   | audio       |
   | output  | audio       |
 
-  | Parameter   | Position  | Default  | Requirement              |
-  | ----------- | --------- | -------- | ------------------------ |
+  | Parameter | Position  | Default  | Requirement              |
+  | --------- | --------- | -------- | ------------------------ |
 
 - Node: **a2a-compressor**<br/>
   Purpose: **audio compressor node**<br/>
@@ -621,14 +729,17 @@ The following nodes process audio chunks only.
   | input   | audio       |
   | output  | audio       |
 
-  | Parameter   | Position  | Default  | Requirement              |
-  | ----------- | --------- | -------- | ------------------------ |
-  | **thresholdDb** | *none* | -18 | `n <= 0 && n >= -60` |
-  | **ratio**       | *none* | 4   | `n >= 1 && n <= 20`  |
-  | **attackMs**    | *none* | 10  | `n >= 0 && n <= 100` |
-  | **releaseMs**   | *none* | 50  | `n >= 0 && n <= 100` |
-  | **kneeDb**      | *none* | 6   | `n >= 0 && n <= 100` |
-  | **makeupDb**    | *none* | 0   | `n >= 0 && n <= 100` |
+  | Parameter       | Position  | Default      | Requirement              |
+  | --------------- | --------- | ------------ | ------------------------ |
+  | **type**        | *none*    | "standalone" | `/^(?:standalone\|sidechain)$/` |
+  | **mode**        | *none*    | "compress"   | `/^(?:compress\|measure\|adjust)$/` |
+  | **bus**         | *none*    | "compressor" | `/^.+$/`             |
+  | **thresholdDb** | *none*    | -23          | `n <= 0 && n >= -100`|
+  | **ratio**       | *none*    | 4.0          | `n >= 1 && n <= 20`  |
+  | **attackMs**    | *none*    | 10           | `n >= 0 && n <= 1000`|
+  | **releaseMs**   | *none*    | 50           | `n >= 0 && n <= 1000`|
+  | **kneeDb**      | *none*    | 6.0          | `n >= 0 && n <= 40`  |
+  | **makeupDb**    | *none*    | 0            | `n >= -24 && n <= 24`|
 
 - Node: **a2a-expander**<br/>
   Purpose: **audio expander node**<br/>
@@ -642,14 +753,15 @@ The following nodes process audio chunks only.
   | input   | audio       |
   | output  | audio       |
 
-  | Parameter   | Position  | Default  | Requirement              |
-  | ----------- | --------- | -------- | ------------------------ |
-  | **thresholdDb** | *none* | -45 | `n <= 0 && n >= -60` |
-  | **ratio**       | *none* | 4   | `n >= 1 && n <= 20`  |
-  | **attackMs**    | *none* | 10  | `n >= 0 && n <= 100` |
-  | **releaseMs**   | *none* | 50  | `n >= 0 && n <= 100` |
-  | **kneeDb**      | *none* | 6   | `n >= 0 && n <= 100` |
-  | **makeupDb**    | *none* | 0   | `n >= 0 && n <= 100` |
+  | Parameter       | Position  | Default  | Requirement           |
+  | --------------- | --------- | -------- | --------------------- |
+  | **thresholdDb** | *none*    | -45      | `n <= 0 && n >= -100` |
+  | **floorDb**     | *none*    | -64      | `n <= 0 && n >= -100` |
+  | **ratio**       | *none*    | 4.0      | `n >= 1 && n <= 20`   |
+  | **attackMs**    | *none*    | 10       | `n >= 0 && n <= 1000` |
+  | **releaseMs**   | *none*    | 50       | `n >= 0 && n <= 1000` |
+  | **kneeDb**      | *none*    | 6.0      | `n >= 0 && n <= 40`   |
+  | **makeupDb**    | *none*    | 0        | `n >= -24 && n <= 24` |
 
 - Node: **a2a-gain**<br/>
   Purpose: **audio gain adjustment node**<br/>
@@ -663,9 +775,9 @@ The following nodes process audio chunks only.
   | input   | audio       |
   | output  | audio       |
 
-  | Parameter   | Position  | Default  | Requirement              |
-  | ----------- | --------- | -------- | ------------------------ |
-  | **db** | *none* | 12 | `n >= -60 && n <= -60` |
+  | Parameter | Position  | Default  | Requirement           |
+  | --------- | --------- | -------- | --------------------- |
+  | **db**    | 0         | 0        | `n >= -60 && n <= 60` |
 
 - Node: **a2a-pitch**<br/>
   Purpose: **audio pitch shifting and time stretching**<br/>
@@ -701,8 +813,9 @@ The following nodes process audio chunks only.
   | input   | audio       |
   | output  | audio       |
 
-  | Parameter   | Position  | Default  | Requirement              |
-  | ----------- | --------- | -------- | ------------------------ |
+  | Parameter   | Position  | Default  | Requirement            |
+  | ----------- | --------- | -------- | ---------------------- |
+  | **segment** | 0         | 50       | `n >= 10 && n <= 1000` |
 
 ### Audio-to-Text Nodes
 
@@ -719,7 +832,7 @@ The following nodes convert audio to text chunks.
 
   | Port    | Payload     |
   | ------- | ----------- |
-  | input   | text        |
+  | input   | audio       |
   | output  | text        |
 
   | Parameter    | Position  | Default  | Requirement        |
@@ -770,9 +883,32 @@ The following nodes convert audio to text chunks.
   | ------------ | --------- | -------- | ------------------ |
   | **key**      | *none*    | env.SPEECHFLOW\_DEEPGRAM\_KEY | *none* |
   | **keyAdm**   | *none*    | env.SPEECHFLOW\_DEEPGRAM\_KEY\_ADM | *none* |
-  | **model**    | 0         | "nova-3" | *none* |
+  | **model**    | 0         | "nova-2" | *none* |
   | **version**  | 1         | "latest" | *none* |
   | **language** | 2         | "multi"  | *none* |
+  | **interim**  | 3         | false    | *none* |
+
+- Node: **a2t-google**<br/>
+  Purpose: **Google Cloud Speech-to-Text conversion**<br/>
+  Example: `a2t-google(language: "en-US")`<br/>
+  Notice: this node requires a Google Cloud API key!
+
+  > This node uses Google Cloud Speech-to-Text to perform Speech-to-Text (S2T)
+  > conversion, i.e., it recognizes speech in the input audio stream and
+  > outputs a corresponding text stream. It supports various languages
+  > and models, including the `latest_long` model for long-form audio.
+
+  | Port    | Payload     |
+  | ------- | ----------- |
+  | input   | audio       |
+  | output  | text        |
+
+  | Parameter    | Position  | Default       | Requirement  |
+  | ------------ | --------- | ------------- | ------------ |
+  | **key**      | *none*    | env.SPEECHFLOW\_GOOGLE\_KEY | *none* |
+  | **model**    | 0         | "latest_long" | *none*       |
+  | **language** | 1         | "en-US"       | *none*       |
+  | **interim**  | 2         | false         | *none*       |
 
 ### Text-to-Text Nodes
 
@@ -783,73 +919,65 @@ The following nodes process text chunks only.
   Example: `t2t-deepl(src: "de", dst: "en")`<br/>
   Notice: this node requires an API key!
 
-  > This node performs translation between English and German languages.
+  > This node performs translation between multiple languages.
 
   | Port    | Payload     |
   | ------- | ----------- |
   | input   | text        |
   | output  | text        |
 
-  | Parameter    | Position  | Default  | Requirement        |
-  | ------------ | --------- | -------- | ------------------ |
-  | **key**      | *none*    | env.SPEECHFLOW\_DEEPL\_KEY | *none* |
-  | **src**      | 0         | "de"     | `/^(?:de\|en)$/` |
-  | **dst**      | 1         | "en"     | `/^(?:de\|en)$/` |
+  | Parameter    | Position  | Default    | Requirement                   |
+  | ------------ | --------- | ---------- | ----------------------------- |
+  | **key**      | *none*    | env.SPEECHFLOW\_DEEPL\_KEY | *none*      |
+  | **src**      | 0         | "de"       | `/^(?:de\|en\|fr\|it)$/`      |
+  | **dst**      | 1         | "en"       | `/^(?:de\|en\|fr\|it)$/`      |
+  | **optimize** | 2         | "latency"  | `/^(?:latency\|quality)$/`    |
 
 - Node: **t2t-amazon**<br/>
   Purpose: **AWS Translate Text-to-Text translation**<br/>
   Example: `t2t-amazon(src: "de", dst: "en")`<br/>
   Notice: this node requires an API key!
 
-  > This node performs translation between English and German languages.
+  > This node performs translation between multiple languages.
 
   | Port    | Payload     |
   | ------- | ----------- |
   | input   | text        |
   | output  | text        |
 
-  | Parameter    | Position  | Default  | Requirement        |
-  | ------------ | --------- | -------- | ------------------ |
-  | **key**      | *none*    | env.SPEECHFLOW\_AMAZON\_KEY | *none* |
+  | Parameter    | Position  | Default  | Requirement                  |
+  | ------------ | --------- | -------- | ---------------------------- |
+  | **key**      | *none*    | env.SPEECHFLOW\_AMAZON\_KEY | *none*     |
   | **secKey**   | *none*    | env.SPEECHFLOW\_AMAZON\_KEY\_SEC | *none* |
-  | **region**   | *none*    | "eu-central-1" | *none* |
-  | **src**      | 0         | "de"     | `/^(?:de\|en)$/` |
-  | **dst**      | 1         | "en"     | `/^(?:de\|en)$/` |
+  | **region**   | *none*    | "eu-central-1" | *none*               |
+  | **src**      | 0         | "de"     | `/^(?:de\|en\|fr\|it)$/`     |
+  | **dst**      | 1         | "en"     | `/^(?:de\|en\|fr\|it)$/`     |
 
-- Node: **t2t-openai**<br/>
-  Purpose: **OpenAI/GPT Text-to-Text translation and spelling correction**<br/>
-  Example: `t2t-openai(src: "de", dst: "en")`<br/>
-  Notice: this node requires an OpenAI API key!
+- Node: **t2t-opus**<br/>
+  Purpose: **OPUS-MT Text-to-Text translation**<br/>
+  Example: `t2t-opus(src: "de", dst: "en")`<br/>
 
   > This node performs translation between English and German languages
-  > in the text stream or (if the source and destination language is
-  > the same) spellchecking of English or German languages in the text
-  > stream. It is based on the remote OpenAI cloud AI service and uses
-  > the GPT-4o-mini LLM.
+  > in the text stream. It is based on the local OPUS-MT translation model.
 
   | Port    | Payload     |
   | ------- | ----------- |
   | input   | text        |
   | output  | text        |
 
-  | Parameter    | Position  | Default  | Requirement        |
-  | ------------ | --------- | -------- | ------------------ |
-  | **api**      | *none*    | "https://api.openai.com" | `/^https?:\/\/.+?:\d+$/` |
+  | Parameter    | Position  | Default  | Requirement      |
+  | ------------ | --------- | -------- | ---------------- |
   | **src**      | 0         | "de"     | `/^(?:de\|en)$/` |
   | **dst**      | 1         | "en"     | `/^(?:de\|en)$/` |
-  | **key**      | *none*    | env.SPEECHFLOW\_OPENAI\_KEY | *none* |
-  | **model**    | *none*    | "gpt-5-mini" | *none* |
 
-- Node: **t2t-ollama**<br/>
-  Purpose: **Ollama/Gemma Text-to-Text translation and spelling correction**<br/>
-  Example: `t2t-ollama(src: "de", dst: "en")`<br/>
-  Notice: this node requires Ollama to be installed!
+- Node: **t2t-google**<br/>
+  Purpose: **Google Cloud Translate Text-to-Text translation**<br/>
+  Example: `t2t-google(src: "de", dst: "en")`<br/>
+  Notice: this node requires a Google Cloud API key and project ID!
 
-  > This node performs translation between English and German languages
-  > in the text stream or (if the source and destination language is
-  > the same) spellchecking of English or German languages in the text
-  > stream. It is based on the local Ollama AI service and uses the
-  > Google Gemma 3 LLM.
+  > This node performs translation between multiple languages
+  > in the text stream using Google Cloud Translate API.
+  > It supports German, English, French, and Italian languages.
 
   | Port    | Payload     |
   | ------- | ----------- |
@@ -858,48 +986,83 @@ The following nodes process text chunks only.
 
   | Parameter    | Position  | Default  | Requirement        |
   | ------------ | --------- | -------- | ------------------ |
-  | **api**      | *none*    | "http://127.0.0.1:11434" | `/^https?:\/\/.+?:\d+$/` |
-  | **model**    | *none*    | "gemma3:4b-it-q4_K_M" | *none* |
-  | **src**      | 0         | "de"     | `/^(?:de\|en)$/` |
-  | **dst**      | 1         | "en"     | `/^(?:de\|en)$/` |
+  | **key**      | *none*    | env.SPEECHFLOW\_GOOGLE\_KEY | *none* |
+  | **src**      | 0         | "de"     | `/^(?:de\|en\|fr\|it)$/` |
+  | **dst**      | 1         | "en"     | `/^(?:de\|en\|fr\|it)$/` |
 
-- Node: **t2t-transformers**<br/>
-  Purpose: **Transformers Text-to-Text translation**<br/>
-  Example: `t2t-transformers(src: "de", dst: "en")`<br/>
+- Node: **t2t-translate**<br/>
+  Purpose: **LLM-based Text-to-Text translation**<br/>
+  Example: `t2t-translate(src: "de", dst: "en")`<br/>
+  Notice: this node requires an LLM provider (Ollama by default, or cloud-based OpenAI/Anthropic/Google, or local HuggingFace Transformers)!
 
   > This node performs translation between English and German languages
-  > in the text stream. It is based on local OPUS or SmolLM3 LLMs.
+  > in the text stream using an LLM service. Multiple LLM providers are
+  > supported: local Ollama (default), local HuggingFace Transformers,
+  > or cloud-based OpenAI, Anthropic, or Google.
 
   | Port    | Payload     |
   | ------- | ----------- |
   | input   | text        |
   | output  | text        |
 
-  | Parameter    | Position  | Default  | Requirement      |
-  | ------------ | --------- | -------- | ---------------- |
-  | **model**    | *none*    | "OPUS"   | `/^(?:OPUS\|SmolLM3)$/` |
-  | **src**      | 0         | "de"     | `/^(?:de\|en)$/` |
-  | **dst**      | 1         | "en"     | `/^(?:de\|en)$/` |
+  | Parameter    | Position  | Default                  | Requirement                              |
+  | ------------ | --------- | ------------------------ | ---------------------------------------- |
+  | **src**      | 0         | "de"                     | `/^(?:de\|en)$/`                         |
+  | **dst**      | 1         | "en"                     | `/^(?:de\|en)$/`                         |
+  | **provider** | *none*    | "ollama"                 | `/^(?:openai\|anthropic\|google\|ollama\|transformers)$/` |
+  | **api**      | *none*    | "http://127.0.0.1:11434" | `/^https?:\/\/.+?(:\d+)?$/`              |
+  | **model**    | *none*    | "gemma3:4b-it-q4\_K\_M"  | *none*                                   |
+  | **key**      | *none*    | ""                       | *none*                                   |
+
+- Node: **t2t-spellcheck**<br/>
+  Purpose: **LLM-based Text-to-Text spellchecking**<br/>
+  Example: `t2t-spellcheck(lang: "en")`<br/>
+  Notice: this node requires an LLM provider (Ollama by default, or cloud-based OpenAI/Anthropic/Google, or local HuggingFace Transformers)!
+
+  > This node performs spellchecking of English or German text using an
+  > LLM service. It corrects spelling mistakes, adds missing punctuation,
+  > but preserves grammar and word choice. Multiple LLM providers are
+  > supported: local Ollama (default), local HuggingFace Transformers,
+  > or cloud-based OpenAI, Anthropic, or Google.
 
-- Node: **t2t-google**<br/>
-  Purpose: **Google Cloud Translate Text-to-Text translation**<br/>
-  Example: `t2t-google(src: "de", dst: "en")`<br/>
-  Notice: this node requires a Google Cloud API key and project ID!
+  | Port    | Payload     |
+  | ------- | ----------- |
+  | input   | text        |
+  | output  | text        |
 
-  > This node performs translation between multiple languages
-  > in the text stream using Google Cloud Translate API.
-  > It supports German, English, French, and Italian languages.
+  | Parameter    | Position  | Default                  | Requirement                              |
+  | ------------ | --------- | ------------------------ | ---------------------------------------- |
+  | **lang**     | 0         | "en"                     | `/^(?:en\|de)$/`                         |
+  | **provider** | *none*    | "ollama"                 | `/^(?:openai\|anthropic\|google\|ollama\|transformers)$/` |
+  | **api**      | *none*    | "http://127.0.0.1:11434" | `/^https?:\/\/.+?(:\d+)?$/`              |
+  | **model**    | *none*    | "gemma3:4b-it-q4\_K\_M"  | *none*                                   |
+  | **key**      | *none*    | ""                       | *none*                                   |
+
+- Node: **t2t-punctuation**<br/>
+  Purpose: **LLM-based punctuation restoration**<br/>
+  Example: `t2t-punctuation(lang: "en")`<br/>
+  Notice: this node requires an LLM provider (Ollama by default, or cloud-based OpenAI/Anthropic/Google, or local HuggingFace Transformers)!
+
+  > This node performs punctuation restoration using an LLM service.
+  > It adds missing punctuation marks (periods, commas, question marks,
+  > exclamation marks, colons, semicolons) and capitalizes the first
+  > letters of sentences. It preserves all original words exactly as they
+  > are without spelling corrections or grammar changes. Multiple LLM
+  > providers are supported: local Ollama (default), local HuggingFace
+  > Transformers, or cloud-based OpenAI, Anthropic, or Google.
 
   | Port    | Payload     |
   | ------- | ----------- |
   | input   | text        |
   | output  | text        |
 
-  | Parameter    | Position  | Default  | Requirement        |
-  | ------------ | --------- | -------- | ------------------ |
-  | **key**      | *none*    | env.SPEECHFLOW\_GOOGLE\_KEY | *none* |
-  | **src**      | 0         | "de"     | `/^(?:de\|en\|fr\|it)$/` |
-  | **dst**      | 1         | "en"     | `/^(?:de\|en\|fr\|it)$/` |
+  | Parameter    | Position  | Default                  | Requirement                              |
+  | ------------ | --------- | ------------------------ | ---------------------------------------- |
+  | **lang**     | 0         | "en"                     | `/^(?:en\|de)$/`                         |
+  | **provider** | *none*    | "ollama"                 | `/^(?:openai\|anthropic\|google\|ollama\|transformers)$/` |
+  | **api**      | *none*    | "http://127.0.0.1:11434" | `/^https?:\/\/.+?(:\d+)?$/`              |
+  | **model**    | *none*    | "gemma3:4b-it-q4\_K\_M"  | *none*                                   |
+  | **key**      | *none*    | ""                       | *none*                                   |
 
 - Node: **t2t-modify**<br/>
   Purpose: **regex-based text modification**<br/>
@@ -919,6 +1082,53 @@ The following nodes process text chunks only.
   | **match**    | 0         | ""       | *required*         |
   | **replace**  | 1         | ""       | *required*         |
 
+- Node: **t2t-profanity**<br/>
+  Purpose: **profanity filtering**<br/>
+  Example: `t2t-profanity(lang: "en", placeholder: "***")`<br/>
+
+  > This node filters profanity from the text stream by detecting bad words
+  > and replacing them with a placeholder. It supports English and German
+  > languages and can either replace with a fixed placeholder or repeat
+  > the placeholder character for each character of the detected word.
+
+  | Port    | Payload     |
+  | ------- | ----------- |
+  | input   | text        |
+  | output  | text        |
+
+  | Parameter       | Position  | Default    | Requirement              |
+  | --------------- | --------- | ---------- | ------------------------ |
+  | **lang**        | *none*    | "en"       | `/^(?:en\|de)$/`         |
+  | **placeholder** | *none*    | "\*\*\*"   | *none*                   |
+  | **mode**        | *none*    | "replace"  | `/^(?:replace\|repeat)$/`|
+
+- Node: **t2t-summary**<br/>
+  Purpose: **LLM-based Text-to-Text summarization**<br/>
+  Example: `t2t-summary(lang: "en", size: 4, trigger: 8)`<br/>
+  Notice: this node requires an LLM provider (Ollama by default, or cloud-based OpenAI/Anthropic/Google, or local HuggingFace Transformers)!
+
+  > This node performs text summarization using an LLM service.
+  > It accumulates incoming text sentences and generates a summary after
+  > a configurable number of sentences (trigger). The summary length is
+  > also configurable (size). It supports English and German languages.
+  > Multiple LLM providers are supported: local Ollama (default), local
+  > HuggingFace Transformers, or cloud-based OpenAI, Anthropic, or Google.
+
+  | Port    | Payload     |
+  | ------- | ----------- |
+  | input   | text        |
+  | output  | text        |
+
+  | Parameter    | Position  | Default                  | Requirement                              |
+  | ------------ | --------- | ------------------------ | ---------------------------------------- |
+  | **provider** | *none*    | "ollama"                 | `/^(?:openai\|anthropic\|google\|ollama\|transformers)$/` |
+  | **api**      | *none*    | "http://127.0.0.1:11434" | `/^https?:\/\/.+?(:\d+)?$/`              |
+  | **model**    | *none*    | "gemma3:4b-it-q4\_K\_M"  | *none*                                   |
+  | **key**      | *none*    | ""                       | *none*                                   |
+  | **lang**     | 0         | "en"                     | `/^(?:en\|de)$/`                         |
+  | **size**     | 1         | 4                        | `1 <= n <= 20`                           |
+  | **trigger**  | 2         | 8                        | `1 <= n <= 100`                          |
+
 - Node: **t2t-sentence**<br/>
   Purpose: **sentence splitting/merging**<br/>
   Example: `t2t-sentence()`<br/>
@@ -977,6 +1187,32 @@ The following nodes process text chunks only.
 
 The following nodes convert text chunks to audio chunks.
 
+- Node: **t2a-openai**<br/>
+  Purpose: **OpenAI Text-to-Speech conversion**<br/>
+  Example: `t2a-openai(voice: "nova", model: "tts-1-hd")`<br/>
+  Notice: this node requires an OpenAI API key!
+
+  > This node uses OpenAI TTS to perform Text-to-Speech (T2S)
+  > conversion, i.e., it converts the input text stream into an output
+  > audio stream. It supports six built-in voices and two models:
+  > `tts-1` for lower latency and `tts-1-hd` for higher quality.
+  > The language is automatically detected from the input text and
+  > supports many languages including German, English, French, Spanish,
+  > Chinese, Japanese, and more (no language parameter needed).
+
+  | Port    | Payload     |
+  | ------- | ----------- |
+  | input   | text        |
+  | output  | audio       |
+
+  | Parameter      | Position  | Default   | Requirement        |
+  | -------------- | --------- | --------- | ------------------ |
+  | **key**        | *none*    | env.SPEECHFLOW\_OPENAI\_KEY | *none* |
+  | **api**        | *none*    | "https://api.openai.com/v1" | `/^https?:\/\/.+/` |
+  | **voice**      | 0         | "alloy"   | `/^(?:alloy\|echo\|fable\|onyx\|nova\|shimmer)$/` |
+  | **model**      | 1         | "tts-1"   | `/^(?:tts-1\|tts-1-hd)$/` |
+  | **speed**      | 2         | 1.0       | `0.25 <= n <= 4.0` |
+
 - Node: **t2a-amazon**<br/>
   Purpose: **Amazon Polly Text-to-Speech conversion**<br/>
   Example: `t2a-amazon(language: "en", voice: "Danielle)`<br/>
@@ -996,7 +1232,7 @@ The following nodes convert text chunks to audio chunks.
   | **key**        | *none*    | env.SPEECHFLOW\_AMAZON\_KEY | *none* |
   | **secKey**     | *none*    | env.SPEECHFLOW\_AMAZON\_KEY\_SEC | *none* |
   | **region**     | *none*    | "eu-central-1" | *none* |
-  | **voice**      | 0         | "Amy"     | `^(?:Amy|Danielle|Joanna|Matthew|Ruth|Stephen|Viki|Daniel)$/` |
+  | **voice**      | 0         | "Amy"     | `/^(?:Amy\|Danielle\|Joanna\|Matthew\|Ruth\|Stephen\|Vicki\|Daniel)$/` |
   | **language**   | 1         | "en"      | `/^(?:de\|en)$/`  |
 
 - Node: **t2a-elevenlabs**<br/>
@@ -1018,11 +1254,34 @@ The following nodes convert text chunks to audio chunks.
   | **key**        | *none*    | env.SPEECHFLOW\_ELEVENLABS\_KEY | *none* |
   | **voice**      | 0         | "Brian"   | `/^(?:Brittney\|Cassidy\|Leonie\|Mark\|Brian)$/` |
   | **language**   | 1         | "de"      | `/^(?:de\|en)$/`  |
-  | **speed**      | 2         | 1.00      | `n >= 0`7 && n <= 1.2` |
+  | **speed**      | 2         | 1.00      | `n >= 0.7 && n <= 1.2` |
   | **stability**  | 3         | 0.5       | `n >= 0.0 && n <= 1.0` |
   | **similarity** | 4         | 0.75      | `n >= 0.0 && n <= 1.0` |
   | **optimize**   | 5         | "latency" | `/^(?:latency\|quality)$/` |
 
+- Node: **t2a-google**<br/>
+  Purpose: **Google Cloud Text-to-Speech conversion**<br/>
+  Example: `t2a-google(voice: "en-US-Neural2-J", language: "en-US")`<br/>
+  Notice: this node requires a Google Cloud API key!
+
+  > This node uses Google Cloud Text-to-Speech to perform Text-to-Speech (T2S)
+  > conversion, i.e., it converts the input text stream into an output
+  > audio stream. It supports various voices and languages with configurable
+  > speaking rate and pitch adjustment.
+
+  | Port    | Payload     |
+  | ------- | ----------- |
+  | input   | text        |
+  | output  | audio       |
+
+  | Parameter    | Position  | Default            | Requirement          |
+  | ------------ | --------- | ------------------ | -------------------- |
+  | **key**      | *none*    | env.SPEECHFLOW\_GOOGLE\_KEY | *none*      |
+  | **voice**    | 0         | "en-US-Neural2-J"  | *none*               |
+  | **language** | 1         | "en-US"            | *none*               |
+  | **speed**    | 2         | 1.0                | `0.25 <= n <= 4.0`   |
+  | **pitch**    | 3         | 0.0                | `-20.0 <= n <= 20.0` |
+
 - Node: **t2a-kokoro**<br/>
   Purpose: **Kokoro Text-to-Speech conversion**<br/>
   Example: `t2a-kokoro(language: "en")`<br/>
@@ -1043,6 +1302,26 @@ The following nodes convert text chunks to audio chunks.
   | **language** | 1         | "en"     | `/^en$/`    |
   | **speed**    | 2         | 1.25     | 1.0...1.30  |
 
+- Node: **t2a-supertonic**<br/>
+  Purpose: **Supertonic Text-to-Speech conversion**<br/>
+  Example: `t2a-supertonic(voice: "M1", speed: 1.40)`<br/>
+
+  > This node uses Supertonic to perform Text-to-Speech (T2S) conversion,
+  > i.e., it converts the input text stream into an output audio stream.
+  > It is intended to generate speech. The ONNX models are automatically
+  > downloaded from HuggingFace on first use. It supports English language only.
+
+  | Port    | Payload     |
+  | ------- | ----------- |
+  | input   | text        |
+  | output  | audio       |
+
+  | Parameter    | Position  | Default  | Requirement |
+  | ------------ | --------- | -------- | ----------- |
+  | **voice**    | 0         | "M1"     | `/^(?:M1\|M2\|F1\|F2)$/` |
+  | **speed**    | 1         | 1.40     | `0.5 <= n <= 2.0` |
+  | **steps**    | 2         | 20       | `1 <= n <= 20` |
+
 ### Any-to-Any Nodes
 
 The following nodes process any type of chunk, i.e., both audio and text chunks.
@@ -1064,8 +1343,8 @@ The following nodes process any type of chunk, i.e., both audio and text chunks.
   | Parameter    | Position  | Default  | Requirement           |
   | ------------ | --------- | -------- | --------------------- |
   | **type**     | 0         | "audio"  | `/^(?:audio\|text)$/` |
-  | **name**     | 1         | "filter" | `/^.+$/` |
-  | **var**      | 2         | ""       | `/^(?:meta:.+\|payload:(?:length\|text)\|time:(?:start\|end))$/` |
+  | **name**     | 1         | "filter" | `/^.+?$/` |
+  | **var**      | 2         | ""       | `/^(?:meta:.+\|payload:(?:length\|text)\|time:(?:start\|end)\|kind\|type)$/` |
   | **op**       | 3         | "=="     | `/^(?:<\|<=\|==\|!=\|~~\|!~\|>=\|>)$/` |
   | **val**      | 4         | ""       | `/^.*$/` |