Start with a question
Describe your goal in natural language. The agent calls data and quantitative tools, keeping the analysis and results in the same conversation.
Open-source AI research terminal
Connect market data, factors, backtests and reports in one conversation. Keep the evidence behind every decision.
Q › Compare Sample A with the benchmark and explore the sources of risk.
Source.fetch_barsPrice window · Dates · CoverageDone
Quant.FactorMomentum · Volatility · RSIDone
Quant.RiskMax drawdown · Risk distributionDone
In this example, the sample has both a higher return and a deeper drawdown.
Consider returns, risk and the data window together to put the result in context.
Q › Compare two watchlists over the same date window.
Common datesAlign dates, weights and currency firstDone
Compare a shared time window, keeping weights and data gaps visible.
Interpret simulated watchlist returns separately from actual account performance.
Q › Save a report of this research so we can continue from here next time.
Research window and data coverage
Factors, backtests and risk metrics
Tool trace and source session
Enable Auto Artifacts in the terminal to generate reports. Browse saved reports with /artifact.
Less switching between tools.
More time thinking about the question.
Describe your goal in natural language. The agent calls data and quantitative tools, keeping the analysis and results in the same conversation.
From momentum and volatility to moving-average strategies, test ideas with factors, backtests and risk metrics. Check the data coverage along the way.
Enable report generation in settings. Use Dashboard to review results and Trajectory to revisit tool calls and the research process.
Choose your model and data sources.
Check the research environment.
/config · /doctor
Build your watchlists.
Keep instruments and questions together.
/portfolio
Ask for analysis in natural language.
Review backtests, risk and evidence.
Q › Analyze this idea
Revisit reports and sessions.
Continue exploring open questions.
/artifact · /resume
Connect models, manage data sources and adjust your workspace with /config in the terminal. Make the tools fit the way you research.
Configuration guideThis is an interactive website demo. Market colors and the Overview toggle update the workspace above. It does not connect to models or read your local configuration.
/config
One market color convention across the workspace
Show market data and watchlists in the sidebar
Adjust animations on this website only
Choose a provider and model ID in the terminal
Choose a reasoning level supported by your model
Save the connection and send a small verification request
Choose a market, data source and sample window
Check a sample response from your connection
Save a verified data connection
A dedicated desktop window for your research, with the runtimes included.
Downloads follow the latest public release. The installer may not yet include the design updates shown on this website.
WhyJ Quant is a quantitative research terminal installed on your computer. This website introduces the product and demonstrates its interface. It does not connect to brokerage accounts or execute trades.
Yes. Use /config to select a model connection and data sources for your markets. Some sources require an API key. AKShare requires a local Python environment and its dependencies. Model and data providers charge for their own services.
Configuration, research sessions and reports are stored in your local workspace. Analysis sends necessary requests to your chosen model and data providers; local storage does not mean all computation is offline. This website only saves demo preferences in your browser.
The project is MIT licensed. Visit GitHub to read the code, report issues or contribute improvements. Available features depend on your installed version and configuration.