EXTENDED COURSE

Jan

Five hands-on lessons take you from installing Jan and downloading your first local model through working fully offline on sensitive data, mixing in cloud models per thread, and exposing a local OpenAI-compatible API at localhost:1337 — an open-source (Apache-2.0), ChatGPT-style assistant that runs 100% on your own machine so no data leaves it.

1Lessons5step-by-step, ~90 min each

Setup

00 Install Jan & download your first local model

Basics

Advanced

2Cheat sheetcopy-ready expressions

3Examples4what people built

Internal tools & ops1

Jan AI-native SME Founder

In the gallery

Meeting transcript cleaner and action-item extractor

Raw auto-transcribed meeting text (filler words, speaker overlaps, repetitions) is pasted into Jan; the model returns a clean summary and a bulleted action-item list with owners and deadlines — sensitive business discussions never leave the local machine.

Clean, actionable meeting notes are ready in under a minute, with confidential strategic discussions processed entirely offline.

Try it yourself

Open Jan, use any instruction model. Prompt: "Clean up this meeting transcript. Remove filler words and repetitions. Then list all action items in format: [owner] — [action] — [deadline]." Paste the raw transcript.

Content & marketing1

Jan Small biz Creator

Course starter

Offline product description writer for a small shop

A shop owner with limited budget uses Jan to draft product listings from bullet-point specs — no subscription, no per-request charge — generating SEO-friendly descriptions for a new product line in one sitting on their home computer.

A full batch of product descriptions is written at zero ongoing cost, with no dependency on cloud AI pricing or availability.

Try it yourself

Open Jan, download a small fast model (e.g. Phi-3 Mini), prompt: "Write a 100-word product description for an e-commerce listing. Tone: friendly and clear. Specs: [paste your bullet points]."

Research & data tools1

Jan Scientist

Course starter

Offline research assistant for fieldwork

A field researcher with no reliable internet installs Jan on a laptop before departure, downloads a capable model, and uses it throughout fieldwork to interpret sensor readings, draft field notes in structured format, and answer methodology questions — entirely offline.

Research work continues productively in remote locations with no connectivity, and all raw observations stay on the device.

Try it yourself

Open Jan, download a capable model (e.g. Llama 3 8B or Gemma 2 9B) from the model hub before you leave. In the chat, paste raw field observations and prompt: "Organise these into a structured field note with date, location, observation, and next steps."

Knowledge & docs1

Jan Creator Scientist

Course starter

Personal knowledge base summariser

A writer or researcher pastes their accumulated notes — drafts, excerpts, references — into Jan's chat and asks it to identify recurring themes, suggest a structure, or extract quotable sentences, with everything processed entirely on their own machine.

A private thinking partner that handles years of personal notes without any of that content going to an external server or training dataset.

Try it yourself

Open Jan, select a mid-size model, paste your notes and ask: "What are the three strongest recurring themes here? List supporting quotes from the text for each."