Telegram Whisper Transcribe
by @xela-io
Standalone Telegram bot for voice message transcription via OpenAI Whisper API. No LLM overhead — audio goes directly to Whisper and text comes back in 2-5 s...
clawhub install telegram-whisper-transcribe📖 About This Skill
name: telegram-whisper-transcribe description: Standalone Telegram bot for voice message transcription via OpenAI Whisper API. No LLM overhead — audio goes directly to Whisper and text comes back in 2-5 seconds. Use when you want a dedicated Telegram bot that only transcribes voice messages, audio files, and video notes without routing through an LLM agent. Supports automatic language detection, runs as a systemd user service, and costs only Whisper API pricing ($0.006/min). env_vars: - name: TELEGRAM_BOT_TOKEN description: Telegram Bot API token from @BotFather required: true - name: OPENAI_API_KEY description: OpenAI API key for Whisper transcription required: true
Telegram Whisper Transcribe
Standalone Telegram bot for fast voice transcription. Bypasses LLM orchestration entirely — Telegram audio goes straight to OpenAI Whisper API and the transcript comes back as a reply.
Why standalone?
OpenClaw routes audio through an LLM agent that then calls Whisper via tool use. That adds 10-30s latency and LLM token costs per transcription. This bot eliminates both — direct API call, 2-5s response time.
Requirements
Setup
1. Install
mkdir -p ~/transcribe-bot
cp {baseDir}/scripts/bot.py ~/transcribe-bot/
python3 -m venv ~/transcribe-bot/venv
~/transcribe-bot/venv/bin/pip install python-telegram-bot openai
2. Configure systemd service
{baseDir}/scripts/install.sh
Or manually:
# Create environment file with secrets (restricted permissions)
cat > ~/transcribe-bot/.env << EOF
TELEGRAM_BOT_TOKEN=
OPENAI_API_KEY=
EOF
chmod 600 ~/transcribe-bot/.envCreate systemd service
cat > ~/.config/systemd/user/transcribe-bot.service << EOF
[Unit]
Description=Telegram Transcribe Bot (Whisper API)
After=network-online.target
Wants=network-online.target[Service]
ExecStart=$HOME/transcribe-bot/venv/bin/python3 $HOME/transcribe-bot/bot.py
Restart=always
RestartSec=5
EnvironmentFile=$HOME/transcribe-bot/.env
[Install]
WantedBy=default.target
EOF
systemctl --user daemon-reload
systemctl --user enable transcribe-bot
systemctl --user start transcribe-bot
3. Verify
systemctl --user status transcribe-bot
Features
Manage
systemctl --user status transcribe-bot
systemctl --user restart transcribe-bot
journalctl --user -u transcribe-bot -f
OpenClaw integration note
If you already have OpenClaw running with a Telegram channel, use a separate Telegram bot token for this service. Do not reuse the same bot token — Telegram only allows one poller per bot.
⚙️ Configuration
1. Install
mkdir -p ~/transcribe-bot
cp {baseDir}/scripts/bot.py ~/transcribe-bot/
python3 -m venv ~/transcribe-bot/venv
~/transcribe-bot/venv/bin/pip install python-telegram-bot openai
2. Configure systemd service
{baseDir}/scripts/install.sh
Or manually:
# Create environment file with secrets (restricted permissions)
cat > ~/transcribe-bot/.env << EOF
TELEGRAM_BOT_TOKEN=
OPENAI_API_KEY=
EOF
chmod 600 ~/transcribe-bot/.envCreate systemd service
cat > ~/.config/systemd/user/transcribe-bot.service << EOF
[Unit]
Description=Telegram Transcribe Bot (Whisper API)
After=network-online.target
Wants=network-online.target[Service]
ExecStart=$HOME/transcribe-bot/venv/bin/python3 $HOME/transcribe-bot/bot.py
Restart=always
RestartSec=5
EnvironmentFile=$HOME/transcribe-bot/.env
[Install]
WantedBy=default.target
EOF
systemctl --user daemon-reload
systemctl --user enable transcribe-bot
systemctl --user start transcribe-bot
3. Verify
systemctl --user status transcribe-bot