Multi Modal Telegram Bot

Inspiration

I got earlier access to Gemini 1.5 pro api. I wanted to test its multi-modal capabilities

What it does

You can upload video, image or send text to it, it will reply your query

How we built it

I used Gemini 1.5 pro file API for it. For bot i used python package "pyTelegramBotAPI"

Challenges we ran into

sometime i got rate limit error from gemini 1.5 pro. At the start of my telegram bot users got a reply from the bot in less time. But now it is taking a longer time to get a response from Gemini.

Accomplishments that we're proud of

I am proud of using Gemini's multi-modal capability in telegram bot

What we learned

I learn how to prompt design to make Gemini output concise & relevant. Also, learn how to use LLM capabilities using APIs.

What's next for Int Bot

I hope more users will use my Telegram Chat Bot "int_gem_bot"

Built With

• amazon-ec2
• gemini1.5pro
• pytelegrambotapi
• python