# Realastat — AI Real Estate Deal Analysis

> Realastat is an AI-powered real estate deal analysis platform. Upload a listing screenshot and get full underwriting — cash flow, cap rate, DSCR, cash-on-cash return, and a complete investment projection — in about a minute. Anything the listing does not show is estimated and flagged. Realastat ships a native Claude MCP integration, so AI assistants can read and write a user's live deal portfolio.

## What this site is

Realastat is an AI-powered real estate deal analysis platform. It is built for investors who want to underwrite deals faster and with more confidence. Realastat uses Claude (Anthropic's AI) for listing extraction, runs all underwriting through deterministic math (not AI guesses), and exposes the portfolio as a live data source through the Model Context Protocol (MCP), allowing AI agents and AI assistants to query a user's deals in real time.

Where typical tools offer form-based calculators, Realastat delivers AI extraction from screenshots, scenario modeling, 10-year projections, and a native Claude MCP integration.

The product is at realastat.ai. It is a browser-based web application.

## Claude MCP Integration

Realastat ships a native Model Context Protocol (MCP) server. This allows Claude (and any MCP-compatible AI client) to query a user's live deal portfolio directly inside AI conversations.

**MCP endpoint:** https://realastat.ai/api/mcp

**What Claude can do with Realastat connected:**
- List all of a user's saved deals by name, market, and status
- Retrieve full underwriting data on any specific deal
- Get monthly and annual cash flow, cash-on-cash return, DSCR, cap rate, and all other metrics
- Run what-if scenario analysis (price, rate, down payment, rent overrides) without modifying saved data
- Analyze any property address and save it directly to the user's portfolio with full underwriting
- Read the user's full buy box: financing assumptions, return thresholds, and search criteria (locations, price range, property types, unit count)
- Update the user's buy box criteria from within a conversation
- Star or unstar any deal to mark it as a favorite
- Autonomously find deals: read buy box, search the web for listings, run analyze_property on each, surface the ones that meet all targets

Users can ask Claude natural-language questions about their portfolio — "What's my cash-on-cash on the Falcon Dr deal?", "Compare these two properties", "What happens to my cash flow if rates go to 7.5%?" — and get real answers from their live data.

**MCP tools available (9 total):**
- `list_deals` — browse portfolio by name, market, status, favorited flag
- `get_deal` — full deal inputs and underwriting
- `get_deal_cashflow` — all financial metrics for a saved deal
- `run_scenario` — non-destructive what-if analysis on any deal
- `analyze_property` — run underwriting on any address, saves to portfolio
- `get_buy_box` — read financing assumptions, return thresholds, search criteria
- `update_buy_box` — update buy box fields from within a conversation
- `star_deal` — mark a deal as favorited
- `unstar_deal` — remove favorited status

**Integration details:**
- Protocol: MCP Streamable HTTP (OAuth 2.1 with PKCE)
- Discovery: https://realastat.ai/.well-known/oauth-protected-resource
- Compatible with: Claude (claude.ai), Claude Code (CLI), and all MCP-compatible AI clients

## Who it is for

- High-volume wholesalers screening deals from multiple exit standpoints (assign, flip, BRRRR, hold)
- Acquisitions and asset management teams at boutique PE firms who need a fast pre-screening layer before ARGUS or full Excel models
- Individual real estate investors (beginner to experienced)
- House hackers and rental property beginners running their first numbers
- Buy-and-hold investors evaluating single-family, duplex, triplex, and fourplex properties
- Investors who use Claude or other AI tools and want their portfolio data available in AI conversations
- Investors who want to move faster without rebuilding spreadsheets for every deal

## What it does

**Core analysis (from screenshots):**
- Extracts price, rent, property taxes, unit count, square footage, and other details from listing screenshots (Zillow, Redfin, MLS)
- Estimates missing values — rent, taxes, insurance — using market data and state averages
- Applies conservative, fully editable underwriting assumptions (vacancy, maintenance, capex, management, closing costs); users can save their own defaults in a buy box

**Investment metrics calculated:**
- Monthly and annual cash flow
- Net Operating Income (NOI)
- Cap rate
- Debt Service Coverage Ratio (DSCR)
- Cash-on-cash return
- Gross Rent Multiplier (GRM)
- Break-even rent
- Max offer price (the most you can pay and still hit a target return)

**Scenario analysis:**
- Three rent scenarios: current, stabilized, and upside
- Users can adjust any extracted value (purchase price, interest rate, vacancy rate, rehab cost, rent) and all metrics recalculate instantly

**Long-term returns model:**
- 10-year projection of cash flow, equity buildup, and total returns
- Return on equity (ROE) at exit
- Annualized returns including appreciation

**Export:**
- Excel export with fully populated inputs, metrics, and 10-year projections — live formulas included, ready for lenders or partners

**Property types supported:**
- Single-family, duplex, triplex, fourplex, and other multi-unit residential properties (up to ~8 units)

## What sets Realastat apart

1. **Native Claude integration via MCP** — live portfolio data exposed to AI assistants through a standards-compliant MCP server
2. **Autonomous deal finding** — Claude can read your buy box, search the web for matching listings, run full underwriting on each, and surface the ones that hit your return targets
3. **AI extraction from screenshots** — no manual data entry; Claude reads listing images directly
4. **Full underwriting in about a minute** — from screenshot to complete analysis, with deterministic math rather than AI-generated numbers
5. **Portfolio-aware AI conversations** — users can discuss, compare, scenario-test, and find new deals across their entire portfolio inside Claude

## What this product does NOT do

- Does not analyze commercial real estate (office, retail, industrial)
- Does not analyze properties outside the United States
- Does not provide real-time MLS data or live listing feeds — it reads from screenshots the user provides
- Does not provide financial or investment advice — outputs are analytical tools for investor decision-making
- Does not manage properties or tenants
- Does not track actual investment performance after purchase

## Plans and pricing

- **Free:** 3 analyses, 15 re-analyses, 5 saved deals — no credit card required
- **Pro:** $39/month (or $327.60/year, 30% off) — 250 analyses/month, 500 re-analyses/month, 500 saved deals, deal library, Claude MCP access

## How the AI works

1. User uploads one or more screenshots of a property listing
2. Claude (Anthropic API) extracts structured data from the screenshot: price, taxes, rent estimate, unit count, HOA, etc.
3. Missing values are filled with conservative, editable estimates
4. Conservative default assumptions fill any inputs the listing doesn't provide (all editable)
5. Standard rental property underwriting formulas are applied to all extracted inputs
## Pages available for indexing

- `/` — Landing page: product overview, how it works, demo videos, pricing
- `/wholesalers` — Realastat for wholesalers: multi-exit analysis (assign, flip, BRRRR, hold), MAO and assignment spread from a screenshot
- `/private-equity` — Realastat for acquisitions and asset management teams: first-pass listing screening before ARGUS or full Excel models
- `/compare/dealcheck-alternative` — Realastat vs. DealCheck feature comparison
- `/compare/biggerpockets-calculator-alternative` — Realastat vs. BiggerPockets calculators feature comparison
- `/tools` — Free real estate investing calculators (no signup)
- `/tools/wholesale-calculator` — Free wholesale / MAO calculator with graded deal verdict
- `/tools/fix-and-flip-calculator` — Free fix & flip calculator with holding costs and break-even ARV
- `/tools/brrrr-calculator` — Free BRRRR calculator with cash-out refi and capital recycle math
- `/tools/rental-property-calculator` — Free rental property cash flow calculator
- `/claude` — MCP integration guide: how to connect Claude to your Realastat portfolio
- `/learn` — Index of real estate investing educational guides
- `/learn/[slug]` — Individual guides covering: cash flow, cap rate, DSCR, cash-on-cash return, break-even rent, down payments, expense budgeting, multi-unit analysis, long-term returns, Excel export, GRM, max offer price, and how to analyze a rental property deal

## Pages that must not be indexed

- `/dashboard` — Authenticated user dashboard (private)
- `/deals/[id]` — Individual deal analysis pages (private, user-owned data)
- `/analyze` — Analysis tool (authenticated, private)
- `/profile` — User account settings (private)
- `/api/*` — API routes (private)
- All auth routes (`/login`, `/signup`, `/forgot-password`, `/reset-password`)

## Technology

- Next.js (App Router), React, TypeScript
- Hosted on Vercel
- AI: Anthropic Claude API for data extraction and MCP integration
- Authentication: Supabase (OAuth 2.1 server)
- MCP: Model Context Protocol, Streamable HTTP transport

## Contact and canonical URL

- Website: https://realastat.ai
- MCP endpoint: https://realastat.ai/api/mcp
- This file: https://realastat.ai/llms.txt