OpenChart Merges AI Agents With Real-Time Trading Charts

Overview

A new open-source project named OpenChart is challenging traditional proprietary trading platforms by merging large language model agents directly into real-time technical analysis environments. Developed by the team at longsurf-ai, the application aims to democratize access to sophisticated quantitative tools, allowing traders to combine code-based charting with autonomous research agents.

The platform distinguishes itself by removing the software quotas that typically restrict advanced indicators on major retail trading sites. Instead, it leverages a custom programming language called Tea, alongside compatibility with major AI providers like Anthropic and Google, to create a more flexible and personalized trading workflow.

Bridging Human Intent and Machine Analysis

OpenChart is designed around the concept of agentic trading, where AI does not merely suggest trades but actively manages the research and monitoring phases of an investment strategy. The developers position the tool not as a product to be monetized through subscriptions, but as a prototype for a future where humans and autonomous agents collaborate to manage assets.

"OpenChart brings AI agents into every step of your investment journey: statistical and technical analysis with code and charts, real-time responses to market events with alerts, opportunity discovery through agentic research, and more."

This philosophy is evident in the platform's core features. Users can describe market conditions in natural language, such as setting a watch for specific volume profiles or technical divergences. The system then generates Tea monitors to evaluate incoming data streams. When a predefined condition is met, the agent can investigate the move and provide source-linked annotations pinned directly to the chart bars.

The platform also supports an interactive loop where users can draw lines or zones on a chart. If price crosses these boundaries, an alert can trigger an agent to research the significance of the move in real time, potentially reducing the lag between market events and human decision-making.

Technical Architecture and the Tea Language

A central component of the OpenChart ecosystem is Tea, a programming language specifically built by the same team for analyzing market data. Inspired by the syntax and utility of TradingView's Pine Script, Tea allows users to write custom indicators and complex conditional logic.

Tea supports functions, loops, state management, and collections, enabling the creation of sophisticated quantitative strategies that evolve as data flows in. This custom runtime is initialized during the standard installation process, ensuring that users have both the charting interface and the logic engine ready to deploy simultaneously.

The software stack is built on modern web technologies, utilizing Bun as the primary runtime and Effect for its underlying concurrency and error-handling foundations. This architecture allows the application to handle the high-frequency data requirements of real-time trading while maintaining stability across different operating systems.

Currently, the desktop application supports macOS (Apple Silicon), with instructions provided for running the software from source using Node.js 24 and npm 11.11.1. The repository structure includes extensive documentation, workflows for continuous integration, and patches to ensure compatibility with various dependencies.

Customization and Community Ecosystem

Unlike closed-source competitors, OpenChart emphasizes user autonomy regarding AI models. Traders are not locked into a single provider's API; they can integrate their existing accounts with Claude, Codex, and Gemini models. This flexibility allows users to leverage their preferred language models for market research and indicator development, potentially reducing costs associated with enterprise AI subscriptions.

The project relies heavily on open-source components, including shadcn/ui, Radix UI, and Base UI for its interface, and Jan for visual design inspiration. The development team has acknowledged contributions from libraries such as robust-orientation and kld-intersections for geometric algorithms, which are critical for accurate charting and volume profile calculations.

Community engagement is facilitated through multiple channels, including a dedicated Discord server, a Reddit community, and public GitHub issues. The repository currently holds 23 stars and 4 forks, indicating early-stage growth within the quantitative trading developer niche.

Licensing and Future Roadmap

OpenChart operates under the OpenChart License, which is based on Apache 2.0 with additional conditions. This licensing choice reflects the project's intent to remain accessible while protecting its specific implementation details. Commercial licensing options are available for entities seeking broader integration rights, with inquiries directed to the project's legal contact.

The desktop application itself is free to download and use. However, users should be aware that AI usage costs are borne by the respective agent provider's plans. Additionally, optional cloud market data services may incur separate fees, though local data sources are supported for those wishing to avoid recurring costs.

The roadmap includes continued development of the Tea compiler and runtime, with invitations for community contributions to the built-in studies and documentation. As the project matures, it aims to establish a standard for how AI agents can be embedded directly into financial visualization tools, moving beyond simple chat interfaces to integrated, actionable intelligence.