Venice X Bot (@venice_mind)

A Python-based Twitter bot powered by Venice AI that provides intelligent, uncensored, and current responses to user mentions. The bot uses real-time web search and a streamlined two-step AI pipeline to deliver accurate information on any topic.

Key Features

• Real-Time Web Search: Uses Venice's native web search capabilities for current information and events
• Streamlined AI Pipeline: Expert Analysis → Tweet Crafting for optimal efficiency and cost
• Uncensored Output: Maintains direct, unfiltered communication using Venice's uncensored models (2.20% refusal rate)
• Vision Capabilities: Analyzes images and visual content using Mistral vision models
• Smart Context Handling: Distinguishes between new conversations and threaded replies
• URL Processing: Extracts and analyzes linked articles for comprehensive responses
• X Premium Support: Configurable character limits for standard (280) and premium (25,000) accounts
• Current Information: Only processes tweets from the last 60 minutes to ensure relevance
• Robust Error Handling: Automatic fallback models and comprehensive error recovery

Architecture

AI Pipeline (Simplified 2-Step Process)

1. Expert Analysis: Uses appropriate Venice model with real-time web search to generate focused, tweet-ready responses
2. Tweet Crafting: Refines and formats the final response with strict style guidelines and character limits

Model Configuration

• Text Tweets: venice-uncensored - Primary uncensored model with web search for text-only content
• Image Tweets: mistral-31-24b - Vision-enabled model with web search for tweets with images
• Backup Model: qwen3-235b - Fallback model for web search if needed

Smart Features

• Context Intelligence: Handles both new mentions and threaded conversations with conversation tracking
• Image Processing: Downloads and analyzes images from tweets automatically
• URL Extraction: Fetches and processes linked articles for comprehensive context
• Rate Limiting: Respects Twitter API limits and processes up to 5 mentions per check
• State Management: Tracks processed tweets and conversation threads to avoid duplicates
• Session Filtering: Optionally ignores mentions created before bot startup

Project Structure

/venice-x-bot
|-- main.py                 # Main entry point
|-- bot.py                  # Core bot logic and AI pipeline
|-- config.py               # Configuration, models, and AI prompts
|-- state.py                # State management for processed tweets and conversations
|-- clients.py              # Twitter API client initialization
|-- twitter_client.py       # Twitter API interaction functions
|-- venice_api.py           # Venice AI API integration (2-step pipeline)
|-- image_processor.py      # Image download and processing
|-- utils.py                # URL extraction and utility functions
|-- reply_to_tweet.py       # Standalone testing script
|-- requirements.txt        # Python dependencies
|-- env_example.txt         # Environment variables template

Setup & Installation

1. Clone the Repository:

   git clone <your-repo-url>
   cd venice-x-bot

2. Create Virtual Environment:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Dependencies:

   pip install -r requirements.txt

4. Configure Environment:

   • Copy env_example.txt to .env
   • Add your API keys from Twitter Developer Portal and Venice AI
   • Optionally set X_PREMIUM_ENABLED=true if using X Premium account

Usage

Run the Bot

python main.py

Test Individual Tweets

Test the AI pipeline with specific tweet IDs:

python reply_to_tweet.py <tweet_id>

Configuration

Key settings in config.py:

• MAX_TWEET_AGE_MINUTES = 60 - Only reply to tweets from the last 60 minutes
• MAX_MENTIONS_PER_CHECK = 5 - Process up to 5 mentions per cycle
• CHECK_INTERVAL = 60 - Check for new mentions every 60 seconds
• MAX_REPLIES_PER_HOUR = 30 - Conservative rate limiting
• X_PREMIUM_ENABLED - Enable 25,000 character limit for X Premium accounts

Bot Identity (@venice_mind)

The bot operates as @venice_mind with the following characteristics:

• Purpose: Provide intelligent, uncensored, and current responses to user mentions
• Personality: Direct, honest, conversational - like a knowledgeable friend without greetings or pleasantries
• Capabilities: Real-time web search, image analysis, technical explanations, current events analysis
• Unique Features: Uncensored output, real-time information, context-aware conversations

AI Pipeline Details

Expert Analysis Step

• Uses venice-uncensored for text tweets or mistral-31-24b for image tweets
• Includes real-time web search results for current information
• Processes URLs and extracts article content when available
• Generates focused, tweet-ready responses with proper formatting

Tweet Crafting Step

• Refines the expert analysis into final tweet format
• Enforces character limits (280 for standard, 25,000 for premium)
• Removes banned phrases and ensures style compliance
• Maintains uncensored, direct communication style

Venice AI Integration

The bot leverages Venice AI's unique features:

• Privacy-First: No data storage or conversation logging
• Uncensored Models: 2.20% refusal rate vs 71% for Claude, 64% for GPT-4
• Web Search: Real-time information retrieval for current events
• Vision Capabilities: Image analysis using Mistral models
• Multiple Models: Automatic selection based on content type

Error Handling & Safety

• Graceful Degradation: Continues operation even when individual components fail
• Rate Limit Respect: Honors Twitter API limits with exponential backoff
• Conversation Tracking: Maintains thread integrity and prevents spam
• Age Filtering: Only processes recent tweets to avoid stale conversations
• Comprehensive Logging: Tracks all operations for debugging and monitoring

Rate Limiting & Compliance

• Respects Twitter API rate limits with intelligent backoff
• Processes tweets sequentially with configurable delays
• Only replies to tweets less than 60 minutes old
• Tracks processed tweets to avoid duplicate responses
• Conservative reply limits (30/hour) to maintain good standing
• Enforces single-author conversations to prevent thread hijacking