DEV Community

Tanvi Sharma
Tanvi Sharma

Posted on

GhostStory AI 👻: Creating Spooky Stories with Local Open-Source AI

Hacktoberfest Open-Source AI Challenge: Ghost in the Machine Submission 👻🌼

GhostStory AI 👻

For Hacktoberfest 2026 Week 4, I built GhostStory AI — a spooky story generator powered by local open-source AI.

👻 What does it do?

  • Creates original ghost stories.
  • Lets users choose a main character.
  • Generates stories in different settings.
  • Offers scary, mysterious, and emotional moods.

🛠️ Tech Stack

  • HTML
  • CSS
  • JavaScript
  • Ollama
  • Qwen 2.5 (1.5B)

🧠 Why Open Innovation Matters

GhostStory AI runs Qwen locally through Ollama. This gives users more control over their prompts and reduces dependence on hosted AI services. Open-weight models make AI experimentation more accessible to beginners and developers.

💻 Source Code

GhostStory AI 👻

A spooky story generator powered by open-source AI.

Features

  • Generate original ghost stories.
  • Choose characters and spooky settings.
  • Select scary, mysterious, or emotional moods.
  • Runs locally using Ollama.

Tech Stack

  • HTML, CSS, JavaScript
  • Ollama
  • Qwen 2.5 (1.5B)

Why Open Innovation?

GhostStory AI uses an open-weight model running locally. This gives users more control over their prompts and stories, reduces dependence on hosted AI services, and makes experimentation accessible.

Run Locally

  1. Install Ollama.
  2. Download qwen2.5:1.5b.
  3. Run python -m http.server 8003 in the project folder.
  4. Open http://localhost:8003.

Demo

A demo video link will be added here.




I enjoyed building a small creative project that brings a little mystery and imagination to AI.

devchallenge #hf26challenge #opensource #ai #hacktoberfest

Top comments (0)