FinceptTerminal

FinceptTerminal is a native C++20 desktop application built with Qt6 and embedded Python, designed as a free, open-source alternative to Bloomberg Terminal. It targets portfolio managers, hedge fund analysts, quantitative researchers, and independent traders who need professional-grade financial intelligence without the $27,000/year cost. The platform solves the problem of inaccessible institutional data by combining real-time market feeds, AI-driven research assistants, and advanced analytics in a single cross-platform desktop app.

Built with Qt6 for high-performance UI rendering and C++20 for speed, FinceptTerminal embeds Python 3.11.9 for analytics and leverages QuantLib for quantitative finance modules. It supports 100+ data connectors including Yahoo Finance, FRED, Kraken, and IMF, and integrates AI agents powered by OpenAI, Anthropic, and Ollama. Deployment options include pre-built installers for Windows, Linux, and macOS, Docker for CI/CD, and manual CMake-based builds with pinned dependencies.

What You Get

• Multi-Asset Analytics - Perform DCF models, portfolio optimization, VaR, Sharpe ratio, and derivatives pricing across equities, fixed income, alternatives, and derivatives using embedded Python and QuantLib.
• AI Agents (37+) - Access AI-powered research assistants modeled after Buffett, Graham, Lynch, Munger, and others; supports local LLMs and multi-provider APIs including OpenAI, Anthropic, Gemini, Groq, and Ollama.
• 100+ Data Connectors - Real-time and historical data from 100+ sources including Polygon, Kraken, Yahoo Finance, FRED, IMF, World Bank, AkShare, and government APIs with optional alternative data like Adanos sentiment.
• Real-Time Trading Engine - Execute paper trades and algo strategies with WebSocket integrations for Kraken and HyperLiquid, plus 16 broker APIs including Zerodha, IBKR, Alpaca, and 5paisa.
• QuantLib Suite - 18 quantitative finance modules for pricing, risk modeling, stochastic processes, volatility surfaces, and fixed income analysis — directly integrated into the platform.
• Visual Node Editor - Build automated data pipelines and trading workflows using a drag-and-drop node-based interface with MCP tool integration for modular AI and data processing.
• AI Quant Lab - Develop and test machine learning models, factor discovery algorithms, and high-frequency trading strategies with built-in ML tooling and data preprocessing.
• Global Intelligence Dashboard - Track maritime shipping, geopolitical events, and satellite-derived market signals to inform macro and geopolitical trading strategies.
• Institutional Charting - Apply 50+ technical indicators (RSI, MACD, Bollinger Bands) with multi-timeframe analysis and candlestick pattern recognition used by professional traders.
• Real-Time News with Sentiment Analysis - Aggregate and analyze financial news from global sources to detect market-moving events and sentiment shifts in real time.

Common Use Cases

• Running a hedge fund research desk - Analysts use FinceptTerminal to pull real-time OHLCV data from 19,000+ instruments, run QuantLib-based risk models, and generate AI-powered equity research reports without Bloomberg subscriptions.
• Building a quantitative trading strategy - Quant researchers leverage the AI Quant Lab and embedded Python to backtest ML models, discover alpha factors, and deploy HFT strategies using Kraken and IBKR APIs.
• Managing a family office portfolio - Portfolio managers use the visual node editor to automate data ingestion from FRED and World Bank, combine it with news sentiment, and visualize portfolio allocations across asset classes.
• Teaching financial markets in academia - Economics professors use the open-source platform to demonstrate macro trends, GDP analysis, and institutional trading workflows without licensing costs.

Under The Hood

Architecture

• The project exhibits a moderately well-defined architecture with a clear separation of concerns, particularly in how it interfaces with diverse data sources through dedicated ‘Wrapper’ classes.
• A ‘Components’ directory suggests a modular design, though the interactions between these components and the core data acquisition layer require further investigation.
• The build process, managed through Dockerfiles, demonstrates a commitment to portability and supports multi-architecture builds.
• Error handling is a focus, with dedicated error classes, but strategic exception management could be improved.

Tech Stack

• The core application is built using modern C++20, leveraging the Dear ImGui framework for the user interface.
• A hybrid approach is employed, embedding Python for scripting and potential extensibility.
• Dependencies are carefully pinned, indicating a focus on reproducibility.
• A robust release management system automatically generates update files with checksums for multiple platforms.

Code Quality

• Testing is prevalent, utilizing pytest with a mix of unit and integration tests, and employing mocking to isolate dependencies.
• Code organization within key components like finagent_core appears structured, with clear separation of concerns.
• Naming conventions are generally consistent and descriptive.
• While error handling is present, its frequency suggests a potential need for more strategic implementation.

What Makes It Unique

• The project demonstrates a sophisticated approach to building autonomous agent systems with a focus on agentic memory isolation and a workflow registry for managing multi-agent interactions.
• The wallet service showcases a unique implementation of Ed25519 verification, including custom base58 encoding/decoding.
• The project’s funding model, clearly defined through a funding.json file, is a commendable approach for open-source sustainability.