Subset is an AI Spreadsheet Editor from Subset Labs, Inc., clearly built for finance professionals. Its goal is to improve spreadsheet productivity for investment bankers, PE/VC investors, equity research analysts, and corporate finance teams. Its core use cases include generating P&Ls from PDFs or CIMs, building DCF models, and updating returns models with natural language. In other words, it is not a general-purpose chatbot, but a vertical AI tool for financial spreadsheets.
Based on the available information, Subset focuses on automating repetitive work in financial analysis: reading CIMs, summarizing materials, transposing financial data, creating P&L, retention, KPI and other metrics, and generating DCFs with real formulas. Its emphasis on “actual spreadsheets with real formulas” is a key selling point, as this is more valuable to finance users than plain-text explanations. Its natural-language modeling capability also seems useful for quickly changing assumptions and updating model structures. However, the underlying AI model, formula accuracy, auditability, version control, and error-checking mechanisms are not disclosed.
The website offers a “Try it free” option, so a free trial is available. However, it does not publicly disclose the trial length, usage quota, feature limits, paid plans, or enterprise pricing. For institutional procurement, the lack of pricing transparency and service-level information may make evaluation more difficult.
The strengths are its clear positioning and coverage of multiple workflows, from CIM review to modeling, investment decision-making, and post-investment updates. It can save time on data transfer, formatting, and building first-draft models. The drawbacks are also clear: some capabilities are still marked as On Roadmap, such as cleaning data room Excel files, fully building returns models, updating benchmarking, and quarterly valuation updates. Its terms also state that it does not guarantee content accuracy, reliability, or continuous service availability, so it cannot replace manual modeling and review.
Subset is best suited to finance professionals who handle large volumes of Excel files, CIMs, PDF financial data, and investment models every day—especially teams that need to quickly build first-draft models, extract operating metrics, and reduce grunt work. It is less suitable for organizations that require Chinese localization, strict compliance-oriented data residency, or already rely heavily on internal modeling systems.
The website does not provide information on access from China, Chinese-language support, or payment methods, so china_access is currently assessed as unknown. Its privacy terms indicate that data is processed in the United States, so financial institutions in China should further evaluate cross-border data and compliance risks. Comparable alternatives include Excel Copilot, Google Sheets Gemini, Rows, Equals, and Causal, or using ChatGPT/Claude together with Excel/Python to build internal workflows.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on subset.so official site.
subset.so is an Unknown AI Apps provider. TG4G tracks its product information, an overall rating of 8.0/10, and a China-accessibility score of Workable. Click "Visit Official Site" to reach subset.so directly.