Watch on TikTok
Watch the Full Breakdown — @smobyday

Google Finance beta is a full redesign of the product — and it has AI built in at a level that makes it genuinely useful for research. Here's what's new and how to use the standout features.

What Google Finance beta actually is

Go to google.com/finance and you'll get the standard interface. Google Finance beta (available in Google Labs) is a rebuilt version with: AI Deep Search (powered by Gemini), integrated prediction market data from Kalshi and Polymarket, live earnings call audio with real-time AI transcription and insights, and a redesigned portfolio and watchlist interface.

To access: go to labs.google.com and enable the Google Finance experiment. Or go directly to google.com/finance/beta while signed into your Google account. Access to Deep Search's higher query limits is available with Google AI Pro or Ultra subscriptions.

Deep Search: the most powerful feature

Deep Search is what makes Finance beta different from any other financial research tool. Instead of returning a simple answer, it issues hundreds of sub-queries simultaneously via Gemini, synthesizes the results, and returns a comprehensive, cited response to complex questions.

Examples of questions Deep Search handles well:

These queries would take hours of manual research. Deep Search returns a structured, cited answer in a few minutes. The citations link to the actual sources, so you can verify the data.

Step 1

Access Google Finance beta

Navigate to google.com/finance/beta and sign in with a Google account. You'll see the redesigned interface with a search bar at the top, a portfolio section, and a watchlist. The Deep Search feature appears as a tab or option in the search results for any stock query.

Step 2

Run a Deep Search query

Search for a stock ticker or topic. In the results, look for the "Deep Search" option (may appear as a button or tab). Enter your complex question. Gemini will indicate it's working — complex queries take 2–5 minutes. The response includes a structured answer with multiple sections and cited sources.

Step 3

Check prediction markets for context

Finance beta integrates Kalshi and Polymarket data for economic questions. When you ask about GDP forecasts, inflation expectations, or election outcomes, you'll see crowd prediction data alongside traditional analyst estimates. This adds a real-time market sentiment layer that traditional financial media doesn't capture.

Step 4

Use alongside Claude for full research workflow

The optimal research workflow: Google Finance Deep Search for data gathering and broad research questions → Claude for interpreting what the data means for a specific investment decision. Deep Search is better at pulling and synthesizing financial data. Claude is better at reasoning through implications, risk factors, and personal portfolio context.

Google Finance research to Claude interpretation workflow
I researched [company/sector] using Google Finance Deep Search and found the following key data: [paste Deep Search summary]. Now help me interpret this data in the context of a potential investment decision. Specifically: (1) what are the 2–3 most significant risk factors suggested by this data, (2) what would change your assessment positively or negatively, (3) how does this fit with a diversified portfolio strategy for a [conservative/moderate/aggressive] investor, and (4) what additional information would be most valuable before making a decision?
Stock research starting prompt
Help me research [company ticker] as a potential investment. I want: a plain-English explanation of what the company actually does and how it makes money, its current valuation context (is it expensive or cheap relative to historical ranges and peers), the main bull case and main bear case as stated by analysts, any significant recent news or events that affect the thesis, and 3 questions I should be able to answer before investing. Use web search to find current information.
smobyday tip

Deep Search takes a few minutes for complex queries — that's not a bug. It's running hundreds of searches and synthesizing them. The output quality is worth the wait. Don't submit a simple price question to Deep Search — save it for the research questions that would otherwise take hours.