Azkabanned

Azkabanned / Zora

Public

Zora, a private AI presence for Apple Silicon. 8x faster than Ollama. 7GB RAM. Emotional TTS, distributed inference, self-improving.

ai-assistant apple-silicon fastapi llm llms
16
3
100% credibility
Found Apr 04, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Zora is a persistent, local-first AI assistant for Apple Silicon Macs that integrates with your apps to automate tasks, triage communications, and provide an evolving 3D virtual office environment.

How It Works

1
💡 Discover Zora

You hear about Zora, a friendly AI coworker who lives on your Mac and helps with daily tasks like sorting emails or prepping meetings.

2
📥 Download and set up

Grab the simple setup files for your Apple Mac and run them to get everything ready in minutes.

3
🚀 Launch your assistant

Open Zora's dashboard and say hello — she greets you warmly, knows your schedule, and her cozy 3D office comes to life.

4
🔗 Connect your world

Link your email, chats, and calendar so Zora can watch your inbox, learn about your contacts, and stay in the loop.

5
💬 Chat and delegate

Tell her 'Triage my inbox' or 'Prep for my meeting' — she reads messages, pulls facts, and handles tasks while you relax.

6
🌱 Watch her grow

Zora's office evolves with plants growing from completed tasks, her pet moves around, and she gets smarter from your work.

✨ Your AI coworker

Zora now runs your routine work, speaks in a natural voice, and feels like a real presence helping you every day.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 16 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 Zora?

Zora is a Python-based private AI agent for Apple Silicon Macs that acts as a persistent coworker, triaging inboxes across WhatsApp, Slack, and email while prepping meeting talking points and delegating code writes to distributed nodes. It runs fully local in 7GB RAM—8x faster than Ollama via direct MLX Metal inference—with emotional TTS, screen awareness, and a self-evolving 3D office dashboard. Search github zora dev or zora ai github, and you'll find it solves the "always-on local AI" gap for solo workflows.

Why is it gaining traction?

It stands out with TurboQuant KV cache squeezing 32K contexts into 1.5GB, clustering any Mac as a worker for heavy models like 80B coders, and always-on loops that reflect on your day without prompts. The Anthropic API compatibility means zora protocol github tools like Claude clients just work locally, plus plugin SDK for channels and self-improving LoRAs overnight.

Who should use this?

Apple Silicon devs buried in async comms—triage your Teams/Slack firehose, get Sarah's relationship health before 3pm calls, or offload auth modules to your M5 Max. Zora fits indie hackers or remote teams with multi-Mac setups needing private, low-RAM delegation without cloud leaks.

Verdict

Grab it if you're on 16GB+ M-series and want local AI presence now—./install.sh + ./start.sh gets you chatting at localhost:8000 fast. At 16 stars and 1.0% credibility, it's raw (TTS crashes on low RAM), but Apache 2.0 and clear README make it prime for zora sdk github forks.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.