Art Argentum is an environment for building, testing and observing investment agents based on AI models and deterministic trading strategies.

The platform combines market data, price history, macroeconomic information and selected market-intelligence sources with tools for research, historical backtesting and paper trading.

In LLM Trading, the external AI model acts primarily as an analytical engine. Art Argentum provides the market environment that applies position rules, money management and market-data logic and then reflects the strategy against real market quotes.

Art Argentum is not a service that connects an AI model from OpenAI, Anthropic, Alphabet or another provider directly to a brokerage account so that the model can autonomously trade real money.

External AI models do not receive direct access to the user's brokerage account and do not send orders to a broker through the platform. The interpretation or indication produced by an agent also is not automatically forwarded to a broker.

The platform is designed for research, analysis, backtesting, paper trading and observing strategy behaviour on real market data. Users can download available live data, signals and strategy outputs through API-based workflows and use them in their own systems at their own discretion.

In LLM Trading, the external AI model performs research and analyses the context provided to it. Its role is to produce an interpretation of the market in line with the user's strategy and prompt.

Art Argentum handles a separate market and trading layer. It combines that interpretation with real market data, applies the configured position and money-management rules, simulates execution inside the agent environment and records positions, PnL and strategy history.

In practical terms, the AI provides research, analysis and interpretation. Art Argentum provides market data, trading rules, execution simulation and performance tracking. The interpretation is not sent directly to a broker.

An LLM Trading Agent uses a large language model during successive strategy iterations. The model analyses the market context provided to it and produces a current interpretation of that context.

That context can include a single instrument, other instruments, macroeconomic data and broader market conditions. This makes it possible to compare assets, analyse intermarket relationships and build systems that infer from wider market behaviour or correlations between instruments.

An Algorithmic Trading Agent works differently. During the building process, the user can give an LLM a selected range of historical market data and ask it to design and test a strategy around that period. The user can also paste pseudocode, provide existing trading logic or simply describe the type of strategy they want to build.

The model can then help translate the idea into explicit algorithmic rules that can be backtested and iteratively refined. Once approved, the strategy is executed by deterministic code and does not require an LLM call on every iteration.

Both approaches can work across multiple instruments, compare markets, incorporate macroeconomic context and use correlations or broader market behaviour. The key difference is that LLM Trading uses the model while the strategy is running, while Algorithmic Trading uses the model mainly to help build the rules that deterministic code later executes.

No. Analyses, interpretations, signals, agent indications, backtests and simulations presented in Art Argentum are provided for informational, analytical and research purposes.

An indication generated by an agent is not investment advice from Art Argentum or from the third-party AI model provider used by the agent. AI models can make mistakes, and their analysis can be incomplete, outdated, incorrectly interpreted or simply wrong.

There is no guarantee that AI research, market interpretation or other outputs are accurate, and no guarantee that they will lead to profitable investment results. Historical backtests and paper-trading results do not guarantee future performance either.

Users remain responsible for evaluating the information and deciding how, or whether, to use it.

Agents can use market charts and price history and, depending on their configuration, additional context such as macroeconomic data, financial intelligence, monitored public sources and other data available in Art Argentum.

The platform can support multi-market workflows across foreign exchange, equities, cryptocurrencies, commodities and market indices, depending on the data available to a specific project.

Public Intelligence is a browsable collection of structured analyses created from monitored public sources, including selected long-form YouTube material.

Its purpose is to make public commentary and research easier to search, compare and use as analytical context.

Public investment agents are read-only reference agents prepared for observation and comparison. Visitors can inspect selected views such as trading activity, PnL, settings, strategy and backtest results without access to private controls or execution functions.

Art Argentum exposes selected data, research outputs, signals and strategy results through API-based workflows. Available live data can therefore be used in external AI agents, analytical systems, internal applications or other software.

Once data is downloaded, the user decides how it is used outside the platform.

For features that use external AI models, pricing is based on the API pricing of the underlying AI provider together with the Art Argentum platform margin. Users do not need to connect or manage their own provider API key because supported model usage is routed and settled through Art Argentum.