Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

FinWorld: An All-in-One Open-Source Platform for End-to-End Financial AI Research and Deployment

Wentao Zhang, Yilei Zhao, Chuqiao Zong, Xinrun Wang, Bo An

TL;DR

FinWorld tackles fragmentation in financial AI by delivering an end-to-end, open-source platform that unifies multimodal data, diverse AI paradigms, and automated experimentation for four core tasks. Its layered architecture, modular components, and AGENT-oriented tooling enable seamless integration of ML/DL/RL, LLMs, and LLM Agents, with GRPO-based RL for LLMs and AgentOrchestra-based multi-agent collaboration. Empirical results across time-series forecasting, trading, portfolio management, and LLM applications demonstrate improved reproducibility, benchmarking, and deployment efficiency, supported by a large-scale dataset and reasoning-focused LLM datasets. This platform promises to accelerate financial AI research and practical deployment by providing standardized data interfaces, evaluation protocols, and automated presentation workflows.

Abstract

Financial AI holds great promise for transforming modern finance, with the potential to support a wide range of tasks such as market forecasting, portfolio management, quantitative trading, and automated analysis. However, existing platforms remain limited in task coverage, lack robust multimodal data integration, and offer insufficient support for the training and deployment of large language models (LLMs). In response to these limitations, we present FinWorld, an all-in-one open-source platform that provides end-to-end support for the entire financial AI workflow, from data acquisition to experimentation and deployment. FinWorld distinguishes itself through native integration of heterogeneous financial data, unified support for diverse AI paradigms, and advanced agent automation, enabling seamless development and deployment. Leveraging data from 2 representative markets, 4 stock pools, and over 800 million financial data points, we conduct comprehensive experiments on 4 key financial AI tasks. These experiments systematically evaluate deep learning and reinforcement learning algorithms, with particular emphasis on RL-based finetuning for LLMs and LLM Agents. The empirical results demonstrate that FinWorld significantly enhances reproducibility, supports transparent benchmarking, and streamlines deployment, thereby providing a strong foundation for future research and real-world applications. Code is available at Github~\footnote{https://github.com/DVampire/FinWorld}.

FinWorld: An All-in-One Open-Source Platform for End-to-End Financial AI Research and Deployment

TL;DR

FinWorld tackles fragmentation in financial AI by delivering an end-to-end, open-source platform that unifies multimodal data, diverse AI paradigms, and automated experimentation for four core tasks. Its layered architecture, modular components, and AGENT-oriented tooling enable seamless integration of ML/DL/RL, LLMs, and LLM Agents, with GRPO-based RL for LLMs and AgentOrchestra-based multi-agent collaboration. Empirical results across time-series forecasting, trading, portfolio management, and LLM applications demonstrate improved reproducibility, benchmarking, and deployment efficiency, supported by a large-scale dataset and reasoning-focused LLM datasets. This platform promises to accelerate financial AI research and practical deployment by providing standardized data interfaces, evaluation protocols, and automated presentation workflows.

Abstract

Financial AI holds great promise for transforming modern finance, with the potential to support a wide range of tasks such as market forecasting, portfolio management, quantitative trading, and automated analysis. However, existing platforms remain limited in task coverage, lack robust multimodal data integration, and offer insufficient support for the training and deployment of large language models (LLMs). In response to these limitations, we present FinWorld, an all-in-one open-source platform that provides end-to-end support for the entire financial AI workflow, from data acquisition to experimentation and deployment. FinWorld distinguishes itself through native integration of heterogeneous financial data, unified support for diverse AI paradigms, and advanced agent automation, enabling seamless development and deployment. Leveraging data from 2 representative markets, 4 stock pools, and over 800 million financial data points, we conduct comprehensive experiments on 4 key financial AI tasks. These experiments systematically evaluate deep learning and reinforcement learning algorithms, with particular emphasis on RL-based finetuning for LLMs and LLM Agents. The empirical results demonstrate that FinWorld significantly enhances reproducibility, supports transparent benchmarking, and streamlines deployment, thereby providing a strong foundation for future research and real-world applications. Code is available at Github~\footnote{https://github.com/DVampire/FinWorld}.

Paper Structure

This paper contains 57 sections, 18 equations, 17 figures, 7 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. ML, DL, and RL in Financial AI
  4. LLMs and LLM Agents in Financial AI
  5. Financial AI Platforms
  6. Preliminaries
  7. Time Series Forecasting Task
  8. Algorithmic Trading Task
  9. Portfolio Management Task
  10. LLMs Applications
  11. Reinforcement Learning for LLMs
  12. FinWorld
  13. Design Principles
  14. Configuration Layer
  15. Dataset Layer
  16. ...and 42 more sections

Figures (17)

  • Figure 1: Overview of FinWorld.
  • Figure 2: Comparison of LLMs on financial reasoning.
  • Figure 3: Comparison of LLMs on trading.
  • Figure 4: Overview of dataset.
  • Figure 5: Overview of double modes of Kline Chart.
  • ...and 12 more figures

Theorems & Definitions (4)

  • Definition 1: Time Series Forecasting
  • Definition 2: Algorithmic Trading
  • Definition 3: Portfolio Management
  • Definition 4: LLMs Applications