Repolode
Submit
anonnerdfrenzy

anonnerdfrenzy/survival-llm

An offline AI assistant with a built-in survival and first-aid library that runs locally and cites the source passage for each answer.

View on GitHubWebsite
Stars
1,523
Forks
138
Open issues
6
Last commit
9 days ago

What is it?

Survival LLM is a desktop app that pairs a local language model with an offline library of emergency, medical, how-to and encyclopedia articles. Questions are answered by a model running through Ollama on your own computer, so nothing is sent to a server and the app keeps working with the internet off. The model searches the library before answering, and each answer lists the articles it read, with the cited passage highlighted.

The app runs on macOS (Apple Silicon) and Windows. Setup downloads an AI model sized to the machine's memory and one of several library packs, with the full library around 8.8 GB. Users can also add their own PDFs, Word documents, EPUB files, web pages and text files, which are searched alongside the library. A memory feature stores short notes about the user, such as household medical details, and the AI's instructions can be edited.

The project publishes its evaluation scripts and results, comparing several models on blackout and first-aid questions with graded accuracy and harmful-answer counts. It also lists known limits, such as models sometimes naming sources they never opened, which the app checks against the articles actually read.

Key features

  • 01Runs a local AI model through Ollama with no internet needed after setup.
  • 02Searches an offline library of medical, emergency and encyclopedia articles.
  • 03Shows the articles each answer used, with the cited passage highlighted.
  • 04Lets users add their own PDFs, Word files, EPUBs, web pages and text files to search.
  • 05Saves short memory notes about the user, editable in a list, and keeps them on the computer.
  • 06Chooses a model based on available memory and offers a reasoning budget and answer length limit.

Use cases

  • Looking up how to store food, purify water or handle a generator during a power outage.
  • Getting first-aid guidance on burns, bleeding or hypothermia when offline.
  • Searching your own household documents, such as a medicine list or emergency plan, alongside reference material.
  • Running a private offline reference assistant on a laptop in areas with unreliable connectivity.

In collections

Private AI knowledge bases and RAGRun LLMs locallyPrivacy & local-first tools

Topics

#offline#local-llm#emergency#first-aid#rag#ollama#privacy

FAQ

What is survival-llm?
An offline AI assistant with a built-in survival and first-aid library that runs locally and cites the source passage for each answer. Survival LLM is a desktop app that pairs a local language model with an offline library of emergency, medical, how-to and encyclopedia articles. Questions are answered by a model running through Ollama on your own computer, so nothing is sent to a server and the app keeps working with the internet off. The model searches the library before answering, and each answer lists the articles it read, with the cited passage highlighted.
Is survival-llm free to use? What license does it have?
survival-llm is open source under the Other license. Check the license terms before using it in commercial products.
Is survival-llm still maintained?
Actively maintained. The last commit to survival-llm was on Oct 29, 2026.
What can I use survival-llm for?
Looking up how to store food, purify water or handle a generator during a power outage. Getting first-aid guidance on burns, bleeding or hypothermia when offline. Searching your own household documents, such as a medicine list or emergency plan, alongside reference material. Running a private offline reference assistant on a laptop in areas with unreliable connectivity.
Where can I get survival-llm?
The source code and installation instructions are on GitHub: https://github.com/anonnerdfrenzy/survival-llm

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects