中文 日本語
home / by job / Local privacy AI

Best Local Privacy AI Agents in 2026

As of Aug 3, 2026, KanonAgent tracks 57 AI agents for Local privacy AI; this page covers the top 8 by real traction, led by anything-llm (64.3k upvotes).
Local privacy AI means running models and agents on your own hardware so data never leaves the device. Recent open-source releases now deliver usable performance for chat, voice, code, and meetings without any backend calls.
Updated 2026-08-03 · 8 products · live data from KanonAgent
1. anything-llm64.3k upvotes
anything-llm gives a self-hosted platform for local-first LLM agents that keep all data under your control.
The local-LLM browser extension runs models offline inside the browser for developers who need private AI tasks.
peerd runs an entire AI agent harness inside the browser with zero backend, suiting real-time private interactions.
4. Osaurus595 upvotes
Osaurus delivers open-source agents that execute 100% locally on Mac for users who want offline privacy.
5. speech-to-speech5.5k upvotes
speech-to-speech builds local voice agents from open-source models for fully offline spoken conversations.
HART OS runs frontier models locally as an open-source AI operating system without datacenter needs.
7. adris.tech10 upvotes
adris.tech packs a full local AI stack into one Windows app for private automation and workflows.
8. Local AGI8 upvotes
Local AGI bundles the local AI stack into a single usable app for anyone avoiding cloud services.

Evolution timeline

The earliest agent for this job entered the KanonAgent index on 2026-07-01 (TryPost.it); the newest entry landed 2026-08-03 (ShoutFlow). Collection and breakout timestamps are written the moment they happen — append-only, never backfilled.
2026-07-01 first indexed TryPost.it
2026-08-03 newest ShoutFlow

How to choose

Match the agent to your platform first: browser extensions for quick private chat, Mac apps for Osaurus or threadfork, Windows for adris.tech. Check model size against your hardware RAM and GPU before committing. Prefer projects with active local inference backends like llama.cpp or MLX to avoid hidden cloud fallbacks. Test offline voice or meeting tools separately if transcription or speech is the main need.

What the data says

Computed from our index over the 8 products on this page; judgement fields are left blank where we cannot read them (methodology).
AutonomyL2 × 4 (4/8 judged)
Prerequisitesopen source × 5 · self-hostable × 8 · bring your own API key × 2
Common integrationsOpenAI × 2
In the index since2026-07-07 — 2026-07-26
Ranked by real traction from our index — not editorial picks, and no paid placement. Every judgement field requires a source quote; where we cannot read it, we leave it blank. Full criteria, thresholds and known limits: methodology.

FAQ

Which local privacy AI runs on Mac without internet?
Osaurus, threadfork, and MinuteFile all run fully offline on Mac hardware.
Can I run voice agents locally for privacy?
speech-to-speech and Voice Studio provide offline voice synthesis and interaction using open models.
What handles local LLM agents in the browser?
The Mozilla Orbit successor extension and peerd both execute agents inside the browser with no server calls.

Explore more