Getting Started
Dashboard Guides
Trading Guides
Reference
Telegram Integration
Need Help?
Join our community for support, updates, and trading discussions.
Assistant-Powered Token Analysis
ScreenerBot integrates with leading Assistant providers to analyze tokens using Large Language Models (LLMs). Get intelligent insights on token viability, risk assessment, scam detection, and optimal entry/exit timing.
What is the Assistant?
Intelligent Token Evaluation
The Assistant uses Large Language Models to evaluate discovered tokens based on market data, metadata, holder information, and historical patterns. Instead of relying solely on numerical thresholds, the Assistant provides contextual analysis that considers multiple factors simultaneously.
This helps identify promising opportunities that might pass numerical filters but have red flags, or spot gems that fall just outside your criteria but show strong fundamental indicators.
Key Capabilities
Intelligent Filtering
The Assistant analyzes tokens during the filtering stage to provide pass/fail recommendations based on liquidity, volume, holder distribution, and metadata quality.
Entry Analysis
Get Assistant recommendations on whether to enter a position based on current market conditions, price action, volume patterns, and risk indicators.
Exit Suggestions
Receive intelligent exit timing suggestions based on market momentum, profit levels, and risk changes in open positions.
Automatic Scam Detection
The Assistant automatically flags and blacklists tokens with scam indicators like suspicious metadata, poor holder distribution, or known rug pull patterns.
Real-time Processing
Fast API providers like Groq deliver analysis in milliseconds, allowing real-time decision making without slowing down your trading bot.
Multi-Provider Support
Choose from 10 AI providers: OpenAI, Anthropic, Google AI (Gemini), Groq, DeepSeek, Together AI, OpenRouter, Mistral, Fireworks, and Ollama. Configure fallback chains for reliability.
Smart Caching
Built-in response caching reduces API costs and improves speed by reusing analysis for identical token data. Configure cache duration and size limits.
Performance Tuning
Adjust concurrent requests, timeouts, retry logic, and other settings to optimize AI performance for your trading volume and requirements.
Supported Assistant Providers
ScreenerBot supports a wide range of Assistant providers to suit different budgets and requirements:
Free Tier Available
Ultra-fast inference with llama-3.1-70b. Best for real-time trading.
Extremely affordable with generous free tier. Great for high-volume analysis.
Large context window, excellent for analyzing complex token data.
Run models locally for complete privacy and no API costs.
Unified API for multiple models with free trial credits.
Premium Providers
GPT-4o-mini offers the best accuracy for complex analysis tasks.
Claude 3.5 Sonnet provides superior reasoning for risk assessment.
Fast and affordable European AI provider with strong performance.
Access to 200+ models including specialized trading-focused LLMs.
Low latency inference optimized for production workloads.
See the Providers Reference for detailed setup instructions and model recommendations for each provider.
Getting Started
Choose Your Provider
Select an Assistant provider based on your budget and requirements. For free usage, Groq is recommended for speed, while DeepSeek offers the highest free quota. For best accuracy, use OpenAI GPT-4o-mini.
Configure API Keys
Obtain API keys from your chosen provider(s) and add them to your ScreenerBot configuration. You can configure multiple providers for redundancy and automatic fallback.
Enable Assistant Features
The Assistant is disabled by default. Enable the features you want: filtering assistance, entry analysis, exit suggestions, and/or automatic blacklisting based on Assistant recommendations.
Test and Monitor
Start with filtering analysis to see Assistant recommendations in action. Monitor the dashboard to see how the Assistant affects your token screening and adjust settings as needed.
Common Use Cases
Enhanced Safety Filtering
Use the Assistant to analyze metadata, social links, and holder patterns to catch scams that pass numerical filters. Automatically blacklist tokens flagged as suspicious.
Smart Entry Timing
Get Assistant recommendations on whether current market conditions are favorable for entry. Avoid FOMO entries and wait for better opportunities identified by the model.
Profit Optimization
Receive exit suggestions based on momentum analysis and market conditions. Take profits at optimal times or hold for larger gains when conditions are favorable.
High-Volume Screening
Process hundreds of new tokens daily with Assistant pre-screening. Let the model filter out obvious scams and low-quality projects before they reach your strategy engine.
Important Considerations
The Assistant is Not Infallible
The Assistant provides recommendations, not guarantees. Always combine Assistant insights with your own research and risk management. Models can make mistakes, especially with novel scam patterns.
Cost Management
API calls have costs (except free tiers). Monitor your usage and set appropriate rate limits. Start with free providers like Groq or DeepSeek before scaling to premium models.
Next Steps
AI Chat
Interactive chat with tool calling. Ask questions, analyze tokens, and execute actions.
Configuration
Learn how to configure Assistant providers, set up API keys, and enable analysis features.
AI Providers
Detailed information on each supported Assistant provider, pricing, and recommended models.
AI Filtering
Use AI to intelligently filter tokens beyond numeric thresholds for better accuracy.