rupeshs

rupeshs / verity

Public

Perplexity style AI answer engine for AI PCs with CPU,GPU and NPU support

llamacpp llm mcp npu ollama
42
3
100% credibility
Found Feb 09, 2026 at 21 stars 2x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Verity is a fully local AI search engine that combines privacy-focused web search with on-device language models to deliver concise, verifiable answers.

How It Works

1
🔍 Discover Verity

You hear about Verity, a smart search helper that works completely on your computer for private, trustworthy answers.

2
🛠️ Prepare your search foundation

You follow a simple guide to set up a private search tool on your machine so Verity can find web info safely.

3
📥 Install with one click

You download Verity and double-click a button to set everything up automatically, feeling excited as it prepares.

4
Choose how to chat
💬
Text chat mode

Jump into simple back-and-forth questions right in a command window.

🌐
Web screen mode

Open a beautiful web page to type and see answers with fancy visuals.

5
❓ Ask your question

You type in what you want to know, and watch the magic as it searches the web privately.

6
⚡ See it work live

You follow along as it finds info, reads pages, thinks deeply, and builds the answer step by step.

✅ Get grounded answers

You receive clear, fact-checked responses with links to sources, all processed safely on your own computer.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 21 to 42 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is verity?

Verity is a Python-based, Perplexity-style AI answer engine that runs entirely locally on AI PCs, pulling real-time web search via self-hosted SearXNG and reasoning with local LLMs for verifiable, concise responses. It delivers a clean web UI, CLI interactive mode, and FastAPI streaming endpoint, optimized for CPU, GPU, or NPU hardware using OpenVINO or Ollama—no cloud dependency means full privacy. Unlike dm-verity GitHub tools or GitHub verity simulators, it's built for perplexity-style search that grounds answers in fetched sources.

Why is it gaining traction?

It stands out as a lightweight perplexity GitHub clone with perplexity-style search that works offline after setup, supporting perplexity GitHub integration vibes without API keys or subscriptions. Developers dig the seamless switch between Ollama, OpenVINO models, or any OpenAI-compatible server, plus bat scripts for one-click Windows installs including perplexity-style wallpapers in the UI flow. The hook? Instant, citation-backed answers on local iron, beating cloud latency for perplexity GitHub access experiments.

Who should use this?

Hardware tinkerers with Intel Core Ultra AI PCs testing NPU acceleration, indie devs prototyping perplexity GitHub connector apps or perplexity-style image prompt engines locally, and privacy hawks ditching cloud search for self-hosted RAG pipelines. Ideal for perplexity GitHub Copilot alternatives in education setups or perplexity GitHub education demos needing offline perplexity-style graphics.

Verdict

Grab it if you're chasing a perplexity-style guide for local AI search—solid docs and easy .bat launches make it dev-friendly despite 22 stars and 1.0% credibility signaling early maturity. Worth a spin for AI PC owners, but watch for broader model support as it grows.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.