Builder’s Edition · Hybrid by default

Talk to your files locally. Private summaries. No uploads.

🔥 Launch Price: $49 — increases to $79 after v1.0

Privacy-first. Franklin runs entirely on your machine — nothing uploaded.

Docker-based FastAPI + compact Web UI. TinyLlama runs on your machine by default; GPT‑4o is opt‑in.

Download Franklin Core and installers from the official download page:

Download Franklin — Download Page
Buy Franklin BE · $49 Download Franklin Core (Free)
Franklin Core is free to download and run locally. Upgrading to Franklin BE unlocks premium features, hybrid GPT-4o mode, and extended plugin support.

Who Franklin BE Is For

  • Developers wanting private local AI tools
  • Homelab users running local-first setups
  • Writers and researchers needing private summaries
  • Privacy-focused users who avoid cloud uploads
  • Builders who want a customizable AI middleware

Core Features

Local TinyLlama engine

Runs on your machine for private, low-latency inference.

Optional GPT-4o hybrid mode

Opt-in cloud fallback for more capable responses when needed.

Private file summaries

Summarize documents without uploads — everything local by default.

Local filesystem tools

List, read and summarize files from a mounted /memory volume.

FastAPI backend + Web UI

Simple backend and small UI for fast local development.

Plugin support

Extend functionality with custom plugins and integrations.

No cloud uploads

By default, nothing leaves your machine.

Fully offline-first design

Works even without network access using local models.

One reproducible Docker stack

Easy to install and run across platforms via Docker Compose.

Franklin vs Cloud AI

Feature Franklin BE ChatGPT / Copilot
Local file access Yes — direct access to mounted /memory No — requires uploads or API bridges
Offline mode Yes (TinyLlama local) No (cloud service)
No subscription fees One-time purchase available Subscription or pay-as-you-go
Full privacy Data stays on your machine Data sent to cloud provider
Extendable plugins Yes — plugin system Limited or provider-specific
One-time purchase Yes (Builder’s Edition) Typically no
Data stays on your machine Always (unless you opt-in cloud) Not by default

Screenshots

Placeholder images — replace with your screenshots.

Screenshot 1 Screenshot 2 Screenshot 3 Screenshot 4

Privacy & Security

  • 100% local processing — models run on your machine
  • No file uploads by default — your data never leaves your device
  • You control your memory directory — mount whatever you want
  • Optional hybrid mode — cloud only if you explicitly enable it
  • No tracking, no analytics — privacy-first by design

Pricing Breakdown — $49

  • Lifetime license
  • All updates included
  • Local-only mode (TinyLlama) included
  • Hybrid mode optional (GPT-4o)
  • Plugin system included
  • Fully private installation — no recurring fees

Simple Install Steps

1. Install Docker

Ensure Docker & Docker Compose are installed on your machine.

2. Clone the repository

git clone https://github.com/dfrankstudioz/franklin-be.git

3. Run the stack

docker compose up -d

4. Open the UI

Open http://localhost:8000 in your browser

FAQ

Does Franklin upload my files?

No. Everything stays on your machine.

Do I need an OpenAI key?

Only if you want optional GPT-4o hybrid mode. Local mode works without it.

Does Franklin run offline?

Yes. TinyLlama runs fully offline.

What OS is supported?

Windows, macOS, and Linux (via Docker).

Can I extend Franklin?

Yes, through the plugin system.

Is there a subscription?

No. It’s a one-time purchase.