Voice input support

README.md

@@ -155,6 +155,17 @@ Windows build requires cross-compilation setup. Check `.github/workflows/` for C
    - **Temperature** — 0.0-2.0 (default: 0.3)
 3. Click **Save Settings**
 
+### Voice Input (Speech-to-Text)
+
+- Start a local transcription server:
+  - `chmod +x scripts/setup_voice_server.sh`
+  - `./scripts/setup_voice_server.sh` (Docker by default; falls back to uvx or venv)
+  - Optional vars: `PORT`, `MODEL`, `DEVICE`, `COMPUTE_TYPE`, `DOCKER_TAG`
+  - Docker resources: `DOCKER_MEMORY` (e.g. `6g`), `DOCKER_CPUS` (e.g. `6`), `DOCKER_SHM_SIZE` (e.g. `2g`)
+  - Example: `DOCKER_MEMORY=6g DOCKER_CPUS=6 DOCKER_SHM_SIZE=2g ./scripts/setup_voice_server.sh`
+- In **Settings → Voice** set `Voice Base URL` (e.g. `http://localhost:8000/v1`), `Model`, optional `API Key`/`Language`
+- Wait for **Connected**, then use the mic button in the prompt input
+
 ### Example Configurations
 
 **Local vLLM:**

scripts/setup_voice_server.sh

@@ -0,0 +1,219 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+echo "[voice] Setting up local transcription server..."
+
+PORT="${PORT:-8000}"
+HOST="${HOST:-0.0.0.0}"
+# IMPORTANT: the Docker image's /v1/models does NOT include `Systran/faster-whisper-large-v3-turbo`,
+# which causes the built-in Gradio UI to crash with AssertionError.
+# This model id is present in /v1/models on `latest-cpu` and works well as a default.
+MODEL="${MODEL:-Systran/faster-whisper-small}"
+DEVICE="${DEVICE:-cpu}"
+COMPUTE_TYPE="${COMPUTE_TYPE:-int8}"
+GIT_URL="${GIT_URL:-https://github.com/fedirz/faster-whisper-server}"
+DOCKER_IMAGE="${DOCKER_IMAGE:-fedirz/faster-whisper-server}"
+# Docker Hub repo has tags like `latest-cpu` / `latest-cuda` (not plain `latest`)
+DOCKER_TAG="${DOCKER_TAG:-latest-cpu}"
+DOCKER_FIRST="${DOCKER_FIRST:-1}"
+DOCKER_DETACH="${DOCKER_DETACH:-1}"
+DOCKER_NAME="${DOCKER_NAME:-localdesk-voice}"
+DOCKER_MEMORY="${DOCKER_MEMORY:-5g}"
+DOCKER_CPUS="${DOCKER_CPUS:-8}"
+DOCKER_SHM_SIZE="${DOCKER_SHM_SIZE:-}"
+
+# Optional: allow passing memory limit as first arg, e.g. ./scripts/setup_voice_server.sh 8g
+if [ -n "${1:-}" ]; then
+  DOCKER_MEMORY="$1"
+fi
+
+port_is_taken() {
+  local port="$1"
+  if command -v lsof >/dev/null 2>&1; then
+    lsof -nP -iTCP:"${port}" -sTCP:LISTEN >/dev/null 2>&1
+    return $?
+  fi
+  if command -v nc >/dev/null 2>&1; then
+    nc -z 127.0.0.1 "${port}" >/dev/null 2>&1
+    return $?
+  fi
+  return 1
+}
+
+find_free_port() {
+  local start_port="$1"
+  local p="$start_port"
+  local i=0
+  while [ $i -lt 50 ]; do
+    if ! port_is_taken "$p"; then
+      echo "$p"
+      return 0
+    fi
+    p=$((p + 1))
+    i=$((i + 1))
+  done
+  return 1
+}
+
+docker_existing_voice_container_for_port() {
+  local port="$1"
+  local image="${DOCKER_IMAGE}:${DOCKER_TAG}"
+  # Match both IPv4/IPv6 published forms by searching for `:${PORT}->8000`
+  docker ps --format '{{.ID}}\t{{.Image}}\t{{.Ports}}' 2>/dev/null \
+    | awk -v img="$image" -v needle=":${port}->8000" '$2==img && index($0, needle)>0 { print $1; exit }'
+}
+
+ensure_brew_pkg() {
+  local pkg="$1"
+  if command -v brew >/dev/null 2>&1; then
+    if ! brew list "$pkg" >/dev/null 2>&1; then
+      echo "[voice] Installing ${pkg} via Homebrew..."
+      brew install "$pkg"
+    fi
+  fi
+}
+
+# ffmpeg is strongly recommended (format support / decoding)
+if ! command -v ffmpeg >/dev/null 2>&1; then
+  if ! command -v brew >/dev/null 2>&1; then
+    echo "[voice] WARNING: ffmpeg not found and brew is unavailable. Install ffmpeg manually."
+  else
+    ensure_brew_pkg ffmpeg
+  fi
+fi
+
+if [ "$DOCKER_FIRST" = "1" ] && command -v docker >/dev/null 2>&1; then
+  echo "[voice] Preferred mode: Docker"
+
+  if port_is_taken "$PORT"; then
+    EXISTING_ID="$(docker_existing_voice_container_for_port "$PORT" || true)"
+    if [ -n "${EXISTING_ID:-}" ]; then
+      echo "[voice] Port ${PORT} is already in use by an existing ${DOCKER_IMAGE}:${DOCKER_TAG} container (${EXISTING_ID})."
+      echo "[voice] Reusing it."
+      echo "[voice] Base URL for LocalDesk: http://localhost:${PORT}/v1"
+      echo "[voice] Tip: open http://localhost:${PORT}/docs"
+      exit 0
+    fi
+
+    FREE_PORT="$(find_free_port "$PORT" || true)"
+    if [ -z "${FREE_PORT:-}" ]; then
+      echo "[voice] ERROR: port ${PORT} is busy and no free port found in the next 50 ports."
+      echo "[voice] Tip: set PORT=8001 (or stop the service using ${PORT})."
+      exit 1
+    fi
+    echo "[voice] Port ${PORT} is busy. Using free port ${FREE_PORT} instead."
+    PORT="$FREE_PORT"
+  fi
+
+  echo "[voice] Starting Docker container ${DOCKER_IMAGE}:${DOCKER_TAG} on http://localhost:${PORT}"
+  echo "[voice] Docker limits: memory=${DOCKER_MEMORY:-unset} cpus=${DOCKER_CPUS:-unset} shm=${DOCKER_SHM_SIZE:-unset}"
+  echo "[voice] Base URL for LocalDesk: http://localhost:${PORT}/v1"
+  echo "[voice] Tip: open http://localhost:${PORT}/docs"
+
+  DOCKER_LIMIT_ARGS=()
+  if [ -n "${DOCKER_MEMORY}" ]; then
+    DOCKER_LIMIT_ARGS+=(--memory "${DOCKER_MEMORY}")
+  fi
+  if [ -n "${DOCKER_CPUS}" ]; then
+    DOCKER_LIMIT_ARGS+=(--cpus "${DOCKER_CPUS}")
+  fi
+  if [ -n "${DOCKER_SHM_SIZE}" ]; then
+    DOCKER_LIMIT_ARGS+=(--shm-size "${DOCKER_SHM_SIZE}")
+  fi
+
+  if [ "${DOCKER_DETACH}" = "1" ]; then
+    # Run in background so it doesn't die when the terminal closes.
+    docker rm -f "${DOCKER_NAME}" >/dev/null 2>&1 || true
+    docker run -d --name "${DOCKER_NAME}" -p "${PORT}:8000" \
+      "${DOCKER_LIMIT_ARGS[@]}" \
+      -e "WHISPER__MODEL=${MODEL}" \
+      -e "WHISPER__INFERENCE_DEVICE=${DEVICE}" \
+      -e "WHISPER__COMPUTE_TYPE=${COMPUTE_TYPE}" \
+      "${DOCKER_IMAGE}:${DOCKER_TAG}" >/dev/null
+    echo "[voice] Container '${DOCKER_NAME}' started in background."
+    echo "[voice] Follow logs: docker logs -f ${DOCKER_NAME}"
+    exit 0
+  fi
+
+  exec docker run --rm -p "${PORT}:8000" \
+    "${DOCKER_LIMIT_ARGS[@]}" \
+    -e "WHISPER__MODEL=${MODEL}" \
+    -e "WHISPER__INFERENCE_DEVICE=${DEVICE}" \
+    -e "WHISPER__COMPUTE_TYPE=${COMPUTE_TYPE}" \
+    "${DOCKER_IMAGE}:${DOCKER_TAG}"
+fi
+
+echo "[voice] Preferred mode: uvx (no global Python installs)."
+echo "[voice] NOTE: faster-whisper-server is often NOT published on PyPI; we install from Git: ${GIT_URL}"
+if command -v uvx >/dev/null 2>&1; then
+  echo "[voice] Starting faster-whisper-server via uvx (from git) on http://localhost:${PORT}"
+  echo "[voice] Base URL for LocalDesk: http://localhost:${PORT}/v1"
+  echo "[voice] Tip: open http://localhost:${PORT}/docs"
+  exec uvx --from "git+${GIT_URL}" faster-whisper-server \
+    --host "$HOST" --port "$PORT" \
+    --model "$MODEL" --device "$DEVICE" --compute_type "$COMPUTE_TYPE"
+fi
+
+if command -v brew >/dev/null 2>&1; then
+  # uv installs uvx as well
+  ensure_brew_pkg uv
+  if command -v uvx >/dev/null 2>&1; then
+    echo "[voice] Starting faster-whisper-server via uvx (from git) on http://localhost:${PORT}"
+    echo "[voice] Base URL for LocalDesk: http://localhost:${PORT}/v1"
+    echo "[voice] Tip: open http://localhost:${PORT}/docs"
+    exec uvx --from "git+${GIT_URL}" faster-whisper-server \
+      --host "$HOST" --port "$PORT" \
+      --model "$MODEL" --device "$DEVICE" --compute_type "$COMPUTE_TYPE"
+  fi
+fi
+
+echo "[voice] uvx is unavailable or failed."
+if command -v docker >/dev/null 2>&1; then
+  echo "[voice] Starting Docker container ${DOCKER_IMAGE}:${DOCKER_TAG} on http://localhost:${PORT}"
+  echo "[voice] Base URL for LocalDesk: http://localhost:${PORT}/v1"
+  echo "[voice] Tip: open http://localhost:${PORT}/docs"
+  exec docker run --rm -p "${PORT}:8000" \
+    -e "WHISPER__MODEL=${MODEL}" \
+    -e "WHISPER__INFERENCE_DEVICE=${DEVICE}" \
+    -e "WHISPER__COMPUTE_TYPE=${COMPUTE_TYPE}" \
+    "${DOCKER_IMAGE}:${DOCKER_TAG}"
+fi
+
+echo "[voice] uvx is unavailable. Falling back to venv + pip."
+echo "[voice] NOTE: faster-whisper-server often requires Python >= 3.10/3.11."
+
+PYTHON_BIN="${PYTHON_BIN:-}"
+if [ -z "$PYTHON_BIN" ]; then
+  if command -v python3.11 >/dev/null 2>&1; then
+    PYTHON_BIN="python3.11"
+  elif command -v python3.10 >/dev/null 2>&1; then
+    PYTHON_BIN="python3.10"
+  elif command -v python3 >/dev/null 2>&1; then
+    PYTHON_BIN="python3"
+  else
+    echo "[voice] ERROR: python3 not found"
+    exit 1
+  fi
+fi
+
+VENV_DIR="${VENV_DIR:-venv_voice}"
+if [ ! -d "$VENV_DIR" ]; then
+  echo "[voice] Creating venv (${PYTHON_BIN}): $VENV_DIR"
+  "$PYTHON_BIN" -m venv "$VENV_DIR"
+fi
+
+echo "[voice] Activating venv..."
+# shellcheck disable=SC1090
+source "$VENV_DIR/bin/activate"
+
+echo "[voice] Upgrading pip..."
+python -m pip install --upgrade pip
+
+echo "[voice] Installing faster-whisper-server..."
+python -m pip install --upgrade "faster-whisper-server @ git+${GIT_URL}"
+
+echo "[voice] Starting server on http://localhost:${PORT}"
+echo "[voice] Base URL for LocalDesk: http://localhost:${PORT}/v1"
+exec faster-whisper-server --host "$HOST" --port "$PORT" --model "$MODEL" --device "$DEVICE" --compute_type "$COMPUTE_TYPE"
+
+