Speech2Note

Repository

---

REPO BY ULTRAGEOPRO

Bot Description

---

The Speech2Note bot enables creating summaries from audio files.

When a user sends a voice message or an audio file containing human speech, the bot:

1. Saves the file on the server.
2. Converts the audio file to text using GigaChat.
3. Saves the text on the server.
4. Creates a summary using GigaChat.
5. Sends the summary back to the user.

If the user is not registered, the bot will prompt them to do so. After registration, their data will be saved in a Supabase table.

No user data is stored permanently on the server.

Database

---

The database is hosted in Supabase.

The table stores only:

• User ID
• User creation date
• User token balance
• User name

Queue

---

All requests are processed in a queue.

Requests in the queue are executed sequentially, with a delay specified in the .env file. The queue also has a maximum request limit.

How to Use

---

First, create a .env file to store all required environment variables.

Example .env file:

TELEGRAM_TOKEN=<bot_token>  
SUPABASE_URL=<supabase_url>  
SUPABASE_KEY=<supabase_key>  
S2T_AUTH_DATA=<s2t_auth_data>  
T2N_AUTH_DATA=<t2n_auth_data>  

SPLIT_TIMEOUT=45  
QUEUE_TIMEOUT=10  
QUEUE_MAX_LEN=20  

• To get the bot token, use Telegram BotFather.
• To obtain an API key for Supabase, refer to Supabase.
• To get an API key for GigaChat, visit GigaChat.

Next, run the bot with the command:

docker compose up -d  

Support

---

Contact email: [email protected]

Telegram: @UltraGeoDev