Forkable financial models you can actually keep. Open, readable, one click into Layerz.
Most financial models are disposable: every deal, every budget starts from a blank sheet, and the logic is tangled up with the data so you can't audit, version or reuse it. The models here are the opposite. Each one is structured, balanced, and reproducible (the balance sheet ties out every year), and each opens straight into Layerz where you can fork it for free and make it your own.
We share the recipe, not the black box. Every model page shows its structure, its conventions and its real numbers.
| Model | What it is | |
|---|---|---|
| ☕ Coffee Shop NYC | 3-year café business plan: P&L, Cash Flow, Balance Sheet that tie out, plus a CFO dashboard |  |
| 📈 SaaS Reporting & Forecast | B2B SaaS operating model 2024-2028: ARR & unit economics, actuals & variance, 3 statements | |
| 🏗️ Infrastructure Project Finance | Reusable project-finance template (worked example: 500 MW offshore wind, 2026-2046): sized senior debt, DSCR, equity IRR, 3 statements that tie out | |
| 🚀 SaaS Scale-up (Series A) | B2B SaaS Series A plan 2026-2030: ARR build, unit economics, path to profit and runway, 3 statements that tie out | |
| 💰 LBO · Returns (Mid-Market Buyout) | Leveraged buyout 2025-2030: sources & uses, sized debt with cash sweep, IRR & MOIC, 3 statements that tie out | |
| 🏦 LBO · Deal & Valuation (Tech M&A) | $43M software buyout 2024-2028: debt tranches as an editable list, actuals/forecast cutover, live checks, football field | |
| 👥 SaaS Cohort Revenue Projection | Monthly cohort engine 2026-2028: churn by tenure (M0-M35), customer base as a stock, MRR bridge that ties out, 3 statements | |
| 💼 Professional Services (Consulting) | Grade pyramid 2026-2028: headcount × capacity × utilisation × rate, WIP and DSO, 3 statements that tie out | |
| 🛒 E-commerce & Retail (Channel Mix) | DTC, marketplace, own retail and wholesale 2026-2028: contribution margin per channel, 40-day cash cycle, 3 statements | |
| ☀️ HoldCo / SPV Portfolio (Rooftop Solar) | 3 project companies under one holding 2026-2045: debt and DSCR per SPV, full cash sweep, investor IRR & MOIC | |
| 📊 Monthly Close (Actuals vs Budget) | Monthly management P&L with a budget base and an actuals branch: native variance and EBITDA budget-to-actual bridge, no second spreadsheet | |
| ⚖️ Valuation (DCF & Multiples) | Company valuation 2025-2030: 5-year unlevered DCF + trading comps (EV/EBITDA, EV/Revenue, P/E) resolved into a football field per share | |
Two LBOs on purpose: Returns is the sponsor view (sources & uses, cash sweep, IRR and MOIC), Deal & Valuation is the advisory view (capital structure you can restructure in one move, price range, actuals cutover).
Models are also composable: reusable modules (a balanced 3-statement skeleton, a multi-loan debt schedule, a payroll block) can be merged into any model from an agent session over MCP. They are building blocks rather than finished models, so they live in Layerz rather than on this shelf.
Want yours here? → CONTRIBUTING.md
- Open a model in Layerz (free account).
- Fork it: you get a model you own, fully editable.
- Edit the drivers and watch every statement recompute. The balance check stays at 0.
No black box, no lock-in. The export to Excel is always clean.
This repo is a GitHub template. If you teach or create around AI and finance, you can have a library exactly like this one, under your own name: your models, on your GitHub, opening into Layerz.
That's the whole point: your credit, your audience, real artifacts your people can fork and keep instead of prompts they'll lose. And you don't start from an empty folder, you inherit the engine: the CLAUDE.md in this repo tells any capable agent how to render a model page straight from your live Layerz model over MCP, and refresh it in one command. Point your agent at it, add your models, done.
See CONTRIBUTING.md for the two ways in: get a model featured here, or spin up your own library.
Hi, I'm Anthony. I've worked hand in hand with dozens of CFOs and finance teams on their models, and kept watching good ones get thrown away after a single deal or budget. So I started keeping mine here, structured and forkable, and I'm building Layerz to make that the normal way to work.
This library is small on purpose. If it gives you one model you can reuse, it did its job. And if it makes you want your own, even better, that is what the template is for.
Shared by @anthomakr · built with Layerz