iBz-04

iBz-04 / quaynor

Public

AI inference library for mobile devices

www.quaynor.site flutter gguf inference-engine llamacpp llm-inference
22
1
100% credibility
Found May 10, 2026 at 22 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

Quaynor enables developers to embed lightweight, offline AI models into Flutter, React Native, Swift, and Python apps with support for chat, tools, embeddings, and multimodal inputs.

How It Works

1
💡 Dream of smart apps

You want to add helpful AI to your mobile or desktop app that works everywhere without internet or privacy worries.

2
📦 Grab Quaynor

Add the simple library to your project and get ready to bring AI on board.

3
🧠 Pick a brain

Choose a small, ready-to-use AI model from the free model library.

4
🚀 Wake it up

Load the model into your app and watch it come alive with one easy step.

5
💬 Start chatting

Ask questions and get instant, natural responses right in your app.

6
🛠️ Add magic tools

Connect helpers for math, lookups, or even seeing pictures to make it smarter.

🎉 AI-powered app ready

Your creation now thinks privately and quickly on phones, tablets, or computers.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 22 to 22 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 quaynor?

Quaynor is a Rust-powered AI inference library that runs GGUF models from Hugging Face on mobile and desktop devices, leveraging Vulkan or Metal for GPU acceleration. It handles chat completions with streaming, embeddings, cross-encoder reranking, and grammar-constrained tool calling—all offline, no API keys needed. Bindings for Python, Flutter, React Native, and Swift deliver a consistent chat API, letting you load models via paths like "huggingface:bartowski/Qwen_Qwen3-0.6B-Q4_K_M.gguf" and query instantly.

Why is it gaining traction?

Unlike heavier LLM inference libraries, quaynor prioritizes mobile speed with a unified API across platforms—swap Python prototypes to Flutter apps without rewriting prompts or samplers. Features like GBNF tool calling and sampler presets (top-k, temperature) make it production-ready for local apps, standing out as a lightweight alternative to cloud-dependent Hugging Face inference or bulkier servers.

Who should use this?

Flutter and React Native devs building on-device chatbots or RAG pipelines on Android/iOS. Swift iOS teams embedding local Qwen or Mistral models for privacy-focused apps. Python scripters prototyping bayesian inference or causal inference workflows offline.

Verdict

Worth prototyping for mobile LLM inference—docs at quaynor.site and pub.dev/npm/pypi packages get you running fast. At 22 stars and 1.0% credibility, it's early-stage with solid tests but unproven at scale; monitor for stability before production.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.