Simple Blackbox.ai API script

completely free (for now) authentication-less llm with file reading capabilities (only text reading, no image recognition) and larger input capablity

https://blackbox.ai

Installation

1. Install python (3.6-3.11)
2. in terminal

git clone https://github.com/Hecker5556/blackbox_api

3. in terminal

cd blackbox_api

4. in terminal

pip install -r requirements.txt

Usage

In terminal:

python blackbox_api.py

this will open a loop where you can text the bot, and stream the response. every response and query for the session will be sent as history.

terminal commands: exit - exits cls/clear - clears the terminal continue - if ai cuts off, u can send a continue request history - prints out current session history getagents - prints out a list of trending agents setagent - prompts to set agent for use with the bot removeagent - removes the currently used agent upload - upload something from a path or a url

In python

from blackbox_api import backbox_api
async def main():
    async for msg in blackbox_api.blackbox(query="how are you"):
        print(msg, end="")
    trending_agents = await blackbox_api.get_trending_agents()
    # for use with trending_agent parameter

Parameters:

• query (str) - what to ask the ai
• history (list[dict]) - history to give to the ai. should be like [{"content": "", "role": "user"/"assistant"}]
• mode (Literal['continue']) - whether to make ai continue from what it last said, required on if no query
• upload (str OR bytes OR BufferedReader) - file to upload to the ai, can be https link, path, bytes or a reader (open(file, 'rb'))
• proxy (str) - proxy to use, can be socks5 or https
• chunk_size (int) - yield text with that size, if None, yield as you receive the response
• trending_agent(str) - trending agent to use, you can get a list from the get_trending_agents