ChatGPT App

The financial data layer
ChatGPT is missing.

Stop ChatGPT from hallucinating stock data. This free connector gives ChatGPT access to 9,900+ US equities, fundamentals, technicals, and SEC filings so you can screen stocks, run analysis, and backtest strategies with real numbers.

https://mcp.shibui.finance/mcp

Free - No API key - Works with all ChatGPT plans (including Free)

9,900+ equities - 31M+ daily prices - 56 indicators - 1962-present - Updated daily

Also works with Claude (native MCP support)

What is Shibui Finance?

Shibui Finance is a free MCP server that gives AI models direct access to 64 years of US stock market data. It covers 31M+ daily price records, quarterly and annual financial statements, 56 technical indicators, daily valuations, and 6.4M SEC filings (via EDGAR) for 9,900+ NYSE and NASDAQ companies. You ask questions in plain English, and ChatGPT queries the database and returns real, verifiable numbers instead of guessing from training data.

Built by Christian Richter in Berlin. Originally built for Claude (which supports MCP natively), now also works with ChatGPT via developer mode apps.

Connect in 2 minutes

Add Shibui Finance as a custom app in ChatGPT's developer mode. The connection persists across conversations.

Works with all ChatGPT plans - including Free

https://mcp.shibui.finance/mcp

Go to chatgpt.com > Settings > Apps, then click Advanced settings.

ChatGPT Settings showing the Apps page with the Create app button under Advanced settings

Toggle Developer mode on. This allows you to add custom MCP connectors.

ChatGPT Advanced settings with Developer mode toggle enabled

Click Create app. Enter the name, paste the URL above, check "I understand and want to continue," then click Create. On the next screen, click Connect.

ChatGPT New App dialog with Shibui Finance name and MCP server URL filled in

Shibui Finance now appears in your apps list. In any new chat, click the + button to attach it to the conversation.

ChatGPT main interface showing Shibui Finance available in the apps menu
1 / 4

For a more detailed walkthrough, see the full setup guide.

Example conversations

Real queries, real results - tested live against 64 years of US market data. Connect and run these yourself in minutes.

You ask

"Which large-cap stocks have an RSI below 30 right now? Show me the sector, RSI, how far they are from their 200-day moving average, and trend strength (ADX). Sort by market cap."

Big, well-known companies in oversold territory - the kind of list contrarians and swing traders build watchlists from. Without stock data access, ChatGPT would guess these numbers or refuse to answer.
ChatGPT returns
output
10 large-caps (mkt cap >$10B) with RSI < 30 right now:

Symbol         Name                Mkt Cap    RSI   vs SMA200   ADX
─────────────  ──────────────────  ─────────  ────  ─────────   ────
TMO.NYSE       Thermo Fisher        $189.7B   28.3      +1.0%   42.7
CRM.NYSE       Salesforce           $178.8B   28.4     -25.8%   43.6
BX.NYSE        Blackstone           $161.6B   27.0     -21.7%   37.8
ACN.NYSE       Accenture            $139.1B   29.1     -19.6%   42.6
BKNG.NASDAQ    Booking Holdings     $134.2B   28.8     -22.9%   39.3
BSX.NYSE       Boston Scientific    $113.1B   28.4     -24.7%   36.3
PANW.NASDAQ    Palo Alto Networks   $113.1B   27.7     -22.8%   33.3
ADBE.NASDAQ    Adobe                $110.5B   29.1     -26.0%   47.9
INTU.NASDAQ    Intuit               $105.9B   20.9     -43.0%   61.8
ADP.NASDAQ     ADP                   $85.8B   25.2     -24.1%   43.3
You ask

"Show me Apple's quarterly revenue and net income over the last 3 years. Include quarter-over-quarter growth rates."

A time-series table with growth rates - instant earnings call prep, no terminal needed. ChatGPT pulls each quarter's stock data from the database rather than approximating from training data.
ChatGPT returns
output
Apple (AAPL) - quarterly revenue & net income, last 3 years:

Quarter     Revenue      Net Income   Rev QoQ    NI QoQ
─────────   ─────────    ──────────   ────────   ──────
2025-12     $143.8B       $42.1B    +40.3%    +53.3%
2025-09     $102.5B       $27.5B     +9.0%    +17.2%
2025-06      $94.0B       $23.4B     -1.4%     -5.4%
2025-03      $95.4B       $24.8B    -23.3%    -31.8%
2024-12     $124.3B       $36.3B    +30.9%   +146.5%
2024-09      $94.9B       $14.7B    +10.7%    -31.3%
2024-06      $85.8B       $21.5B     -5.5%     -9.3%
2024-03      $90.8B       $23.6B    -24.1%    -30.3%
2023-12     $119.6B       $33.9B    +33.6%    +47.7%
2023-09      $89.5B       $23.0B     +9.4%    +15.5%
2023-06      $81.8B       $19.9B    -13.7%    -17.7%
You ask

"Compare Microsoft, Apple, Google, Amazon, and Meta on P/E ratio, profit margin, ROE, and revenue growth. Who's the best value right now?"

A side-by-side table of the Mag 5 with a reasoned take on relative value. ChatGPT formats the stock data as a ranked comparison - the kind of peer analysis people pay for on Seeking Alpha.
ChatGPT returns
output
Big Tech head-to-head - P/E, margins, ROE, revenue growth (Feb 2026):

Symbol   Name        Mkt Cap      P/E    Margin    ROE     Rev Growth
───────  ──────────  ──────────   ─────  ──────    ──────  ──────────
GOOGL    Alphabet     $3,810B     29.2   32.8%     35.7%    +18.0%
AAPL     Apple        $3,759B     32.4   27.0%    152.0%    +15.7%
MSFT     Microsoft    $3,005B     25.3   39.0%     34.4%    +16.7%
AMZN     Amazon       $2,134B     27.7   10.8%     22.3%    +13.6%
META     Meta         $1,673B     28.2   30.1%     30.2%    +23.8%

MSFT has the lowest P/E at 25.3 and highest profit margins at 39%.
AAPL leads on ROE (152%) driven by aggressive buybacks.
META leads on revenue growth at +23.8% YoY.
You ask

"Over the last 5 years, what happened when large-cap stocks had an RSI below 30? What was the average return 30 days later?"

A statistical backtest across 15,000+ stock data points - the kind of hard evidence most investors never bother to check. ChatGPT scans 5 years of daily prices and computes win rates, not approximations.
ChatGPT returns
output
RSI < 30 backtest - large-caps (mkt cap >$10B), last 5 years:

Metric                    Result
──────────────────────    ──────────
Total signals              15,329
Win rate                  57.6%
Avg 30-day return         +1.75%
Median 30-day return      +1.78%
Avg winning trade         +9.96%
Avg losing trade          -9.40%
Best return              +541.0%
Worst return              -85.1%

57.6% win rate across 15,329 signals - measured, not adjusted for
survivorship or delisting. Directional evidence, not a trade signal.
You ask

"Which large companies disclosed an executive or board change with the SEC in the last 45 days?"

A live feed of C-suite and boardroom changes at big companies - pulled straight from 8-K filings, each linking to the source document on EDGAR. ChatGPT filters 6.4M SEC filing records by form type and date without you writing SQL.
ChatGPT returns
output
8-K filings reporting an executive/board change (item 5.02) - last 45 days, mkt cap >$20B:

Symbol        Name          Mkt Cap   Filed
────────────  ────────────  ───────   ──────────
NVDA.NASDAQ   NVIDIA        $5,131B   2026-05-08
AAPL.NASDAQ   Apple         $4,581B   2026-04-20
MSFT.NASDAQ   Microsoft     $3,343B   2026-05-14
ORCL.NYSE     Oracle          $649B   2026-05-12
XOM.NYSE      Exxon Mobil     $605B   2026-05-04
INTC.NASDAQ   Intel           $573B   2026-04-24
MA.NYSE       Mastercard      $437B   2026-05-07
CVX.NYSE      Chevron         $364B   2026-05-29
C.NYSE        Citigroup       $220B   2026-05-21
VZ.NYSE       Verizon         $202B   2026-05-28

Each row links to the filing on sec.gov. 8-Ks are filed by the company itself.
See 20 real-world analyses by investor type →

ChatGPT guides

How to Get Stock Data into ChatGPT

Step-by-step setup. No code, no API keys.

Stop ChatGPT from Hallucinating Financial Data

Why ChatGPT invents numbers and the structural fix.

Using Claude instead? See the Claude setup guide and all guides and comparisons.

What Shibui does not cover

US equities only (NYSE + NASDAQ). No international markets, options, crypto, or commodities. End-of-day data only, no real-time or intraday prices. Data comes from tier-3 providers, not Bloomberg or Refinitiv. Good for screening and analysis, not institutional-grade for compliance or trade execution. See the data sources page for full coverage details.

Frequently asked questions

Does Shibui Finance work with ChatGPT?

Yes. Shibui Finance works with ChatGPT via MCP developer mode apps. It gives ChatGPT access to 64 years of US stock market data: 31M+ daily prices, quarterly financials, daily valuations, and 56 technical indicators for nearly 10,000 companies. Setup takes about 2 minutes.

How do I connect a stock database to ChatGPT?

Add Shibui Finance as a developer mode app in ChatGPT using the MCP server URL (https://mcp.shibui.finance/mcp). The connection persists across conversations. No API keys, no code, no subscription. Works with all ChatGPT plans, including Free. See the step-by-step setup guide.

Is Shibui Finance free for ChatGPT users?

Yes. Shibui Finance is completely free: no API keys, no subscription, no per-query costs. It works with all ChatGPT plans, including Free, via developer mode. The data is the same as the Claude connection: 9,900+ US equities with 64 years of history.

What stock data can I access in ChatGPT with Shibui?

Shibui gives ChatGPT access to daily OHLCV prices from 1962, quarterly and annual financial statements, daily valuations (P/E, P/B, market cap, EV), 56 pre-calculated technical indicators (RSI, MACD, SMA, Bollinger Bands), quarterly earnings with surprise data, and 6.4M SEC EDGAR filing metadata records, covering nearly 10,000 US-listed securities.

Does Shibui work with ChatGPT Plus?

Yes. Shibui Finance works with all ChatGPT plans, including Free. You add it as a custom app via developer mode. Shibui itself is free regardless of which ChatGPT plan you use.

Connect Shibui to ChatGPT

Shibui is free. Connect it to ChatGPT and get financial data from a verified database instead of the model's training data.

Set up now