Discord Crypto Analyzer

A powerful Discord bot for real-time cryptocurrency market analysis using advanced AI models.

Try It Live

Join our Discord server to see the bot in action and test features in real time:

👉 Discord Invitation Link

Channel: #ai-analysis

Overview

Discord Crypto Analyzer provides real-time cryptocurrency analysis directly to your Discord server. Using advanced AI models, it generates detailed reports, forecasts, and detects key market patterns—perfect for traders and enthusiasts seeking instant, data-driven insights.

Features

Real-time Market Analysis: Analyze any cryptocurrency trading pair with a simple command
AI-Powered Insights: Utilizes advanced AI models to interpret complex market data
Visual Chart Analysis: Automatically generates candlestick charts that AI models analyze for enhanced pattern recognition and market insights
Technical Indicators: Comprehensive analysis using RSI, MACD, Bollinger Bands, Supertrend, and many more.
Multi-Timeframe Analysis: Examines data across multiple timeframes (1h, 2h, 4h, 6h, 8h, 12h, 1d).
Flexible Timeframe Selection: Analyze markets on your preferred timeframe via Discord commands or config
Support & Resistance Levels: Identifies key price levels.
Price Targets: Provides short-term and medium-term price targets
Sentiment Analysis: Incorporates Fear & Greed index for market sentiment
Detailed Reports: Generates in-depth HTML reports for comprehensive analysis
Anti-Spam Protection: Built-in measures to prevent channel spam and abuse
Context-Aware Analysis: Uses RAG (Retrieval-Augmented Generation) for enhanced market context
Multi-Language Support: Analyze markets in multiple languages
Auto-Categorization: Automatically identifies and categorizes crypto assets
Multi-Exchange Support: Fetches data from multiple exchanges including Binance, KuCoin, Gate.io, MEXC, and Hyperliquid
Divergence Detection: Identifies bullish and bearish divergences across multiple indicators
Streaming Analysis: Real-time streaming responses when using local AI models for faster feedback
Automatic Cleanup: Messages and files automatically expire and are cleaned up after a configurable period
Fault-Tolerant Design: Multiple AI model fallbacks for continuous operation
Smart Caching: Optimized data retrieval with minimal API calls
Educational Resources: Automated linking to indicator explanations and definitions
Pattern Recognition: Detects common chart patterns like head and shoulders, wedges, etc.
Market Metrics: Calculates key metrics across various timeframes
Weekly Macro Trends: Analyzes 200-week SMA for institutional-grade macro trend analysis with cycle phase identification

Installation

Prerequisites

Python 3.13 or higher
Discord Bot Token
OpenRouter API Key or Local LM Studio setup
CryptoCompare API Key (for market data)
Google Studio API Key (optional, for Gemini models via official Google GenAI SDK)

Steps

These installation instructions assume you have a local copy of the repository (clone or download). If you're using a fork or mirror, follow the same steps below to create a virtual environment and install dependencies.

Clone or download the repository to your local machine (or use your fork).
(Recommended) Create a Python virtual environment:

python -m venv .venv

Activate the virtual environment:

On Windows (PowerShell):
```
.venv\Scripts\Activate.ps1
```
On Windows (cmd):
```
.venv\Scripts\activate.bat
```
On macOS/Linux:
```
source .venv/bin/activate
```

Install the required dependencies:

pip install -r requirements.txt

Create configuration files:
- Copy keys.env.example to keys.env
- Set your API keys and Discord settings in the keys.env file
- Adjust other settings in config/config.ini as needed
Run the bot:

python start.py

Configuration

The configuration is split across two files:

`keys.env`

Contains sensitive information that should not be committed to version control:

BOT_TOKEN_DISCORD: Your Discord bot token
OPENROUTER_API_KEY: API key for OpenRouter AI services
GOOGLE_STUDIO_API_KEY: API key for Google Studio (Gemini) models
CRYPTOCOMPARE_API_KEY: API key for CryptoCompare data
GUILD_ID_DISCORD: ID of your Discord server
MAIN_CHANNEL_ID: Main channel for bot communication
TEMPORARY_CHANNEL_ID_DISCORD: Channel for temporary file uploads

`config/config.ini`

Contains non-sensitive configuration that can be safely committed.

Timeframe Configuration

The bot supports flexible timeframe analysis from 1 hour to 1 day:

Supported Timeframes: 1h, 2h, 4h, 6h, 8h, 12h, 1d

Setting Default Timeframe

In config/config.ini:

[general]
timeframe = 1h  # Change to 4h, 1d, etc.

This sets the default timeframe used when users don't specify one in their command.

Per-Analysis Timeframe Override

Users can specify timeframe directly in Discord commands:

!analyze BTC/USDT          → Uses config default (1h)
!analyze BTC/USDT 4h       → Analyzes on 4-hour timeframe
!analyze BTC/USDT 1d       → Analyzes on daily timeframe
!analyze BTC/USDT 4h Polish → 4-hour timeframe in Polish

How It Works:

Dynamic Calculations: All technical indicators, period calculations, and candle progress tracking automatically adapt to the selected timeframe
Smart Data Fetching: System automatically adjusts candle limits to maintain ~30 days of historical data across all timeframes
Real-time Flexibility: Users don't need to wait for bot restart to try different timeframes

Choosing Your Timeframe:

1h-2h: Short-term trading, high granularity, more frequent signals
4h-6h: Swing trading, balanced view, medium-term trends
8h-12h: Longer swing positions, reduced noise
1d: Position trading, macro trends, most stable signals

How AI Provider Selection Works

Use the provider setting in the [ai_providers] section of config/config.ini to select the AI provider. This section explains the available options and exact runtime behavior.

provider (string) — valid values and behavior:
- "local": Use LM Studio only (local models). LM Studio is the only provider used; streaming is supported when LM Studio is available. If LM Studio fails, the request will fail — there is no automatic fallback.
- "googleai": Use Google AI Studio only (Gemini models). Google is the single provider; no automatic fallbacks to other providers on failure.
- "openrouter": Use OpenRouter only. OpenRouter is the single provider; no automatic fallbacks to other providers on failure.
- "all": Use the fallback chain. In this mode the runtime will try providers in this order until one returns a valid response:
  1. Google AI Studio
  2. LM Studio (local)
  3. OpenRouter
When provider is set to "all", the system preserves the original fallback behavior. For single-provider settings ("local", "googleai", "openrouter") there is no cross-provider fallback.
lm_studio_base_url: URL for LM Studio (default: "http://localhost:1234/v1")
lm_studio_model: Model identifier for LM Studio (default: "local-model")
OPENROUTER_BASE_URL: Base URL for OpenRouter API (default: "https://openrouter.ai/api/v1")
OPENROUTER_BASE_MODEL: Default OpenRouter model identifier (e.g. "google/gemini-2.5-flash")
OPENROUTER_FALLBACK_MODEL: Fallback model identifier for OpenRouter
GOOGLE_STUDIO_MODEL: Default Google model identifier (e.g. "gemini-2.5-flash")
TIMEFRAME: Default timeframe for analysis (default: "1h")
LOG_DIR: Directory for log files (default: "logs")
LOGGER_DEBUG: Enable debug logging
ANALYSIS_COOLDOWN_COIN: Time between analyses of the same coin (in seconds)
ANALYSIS_COOLDOWN_USER: Time between user requests (in seconds)
RAG_UPDATE_INTERVAL_HOURS: How often to update the RAG (Retrieval-Augmented Generation) system
SUPPORTED_LANGUAGES: Dictionary of supported languages for analysis output

Usage

Commands

!analyze <symbol> [timeframe] [language]: Analyze a trading pair with optional timeframe and language
- Basic: !analyze BTC/USDT (uses config default timeframe, English)
- With timeframe: !analyze BTC/USDT 4h (4-hour timeframe, English)
- With language: !analyze BTC/USDT Polish (config default timeframe, Polish)
- Full: !analyze BTC/USDT 1d Polish (daily timeframe, Polish)
!cleanup: Force cleanup of expired messages (owner only)
!unmute <user>: Unmute a user that was muted for spamming

Supported Timeframes: 1h, 2h, 4h, 6h, 8h, 12h, 1d

Admin: Custom AI Provider & Model Selection

Administrators can override the default AI provider and model for analysis. This is useful for testing different models or using specific providers:

!analyze <symbol> [timeframe] [language] <provider> <model>

Available Providers:

googleai — Google AI Studio (Gemini models)
openrouter — OpenRouter (access to 500+ models: GPT, Claude, Gemini, DeepSeek, etc.)
local — LM Studio (local models)
all — Fallback chain (tries: Google AI → LM Studio → OpenRouter)

Admin Command Examples:

# Use specific Gemini model
!analyze BTC/USDT googleai gemini-2.5-pro

# Use OpenRouter with GPT model
!analyze BTC/USDT openrouter openai

DiscordCryptoAnalyzer

Install / Use

README