In one sentence
NashNova isn't a chattier ChatGPT — it's a research system built for investing. Give a general-purpose LLM web search and it can find "the latest" news, but its sources are still mostly public web pages. NashNova plugs into proprietary structured financial data and integrates it around investment scenarios.
Overall Score Comparison
NashNova leads the runner-up by 14.8 points, and GPT + WebSearch by 23.3. It wins on overall score in 95 of 100 questions.
Where the Gap Shows Up: 7 Dimensions
The three foundational dimensions are data grounding, source traceability, and factual precision, with gaps of roughly 11–12 points. The difference isn't just whether a source can be found — it's whether earnings, research reports, filings, market data, and macro variables can be pulled into a single judgment.
The four higher-order dimensions are reasoning quality, actionability, completeness, and timeliness, with gaps of 13–20 points. On reasoning quality, actionability, and completeness, NashNova leads in 96 of 100 questions each.
5 Market Categories
Macro
NashNova scores 81.7, leading Claude by 19.1 points.
Asset Classes
NashNova scores 79.0, leading Claude by 23.2 points.
U.S. Equities
NashNova scores 81.5, leading Claude by 14.6 points.
A-Shares & Hong Kong Stocks
NashNova still leads by 8.6 points on A-shares and 10.7 on Hong Kong stocks — and it's also an area we're continuing to strengthen.
The Takeaway
A general-purpose LLM is more like a well-read assistant you can talk investing with — good at explaining concepts, organizing viewpoints, and doing broad-strokes analysis.
NashNova takes a different path: it turns research, tracking, and review into a research system that runs continuously. Its edge isn't better conversation — it's better research.
Use a research system, not one-off QA
Hand your holdings, themes, and risk variables to NashNova for continuous tracking.
Start asking →