Parlor

Parlor: On-device multimodal AI application that runs real-time voice and video conversations.

Parlor is an on-device, real-time multimodal artificial intelligence application designed for natural voice and video conversations. It operates locally using Gemma 4 models running through llama.cpp, paired with Silero VAD and Pipecat smart-turn-v3 for hands-free turn-taking. Users can speak and share camera frames while the system streams responses sentence by sentence with barge-in support. Optional background research routes tasks to OpenAI-compatible endpoints. It supports live translation, silent scribe mode, and built-in timers, making it suitable for local voice interactions.

Parlor at a glance

Best for: Software engineer
Runs on: macOS, Linux, Web, self-hosted, open source
Pricing: Free and open source under Apache 2.0 license
Released: April 2026
Category: AI chat assistant
Website: github.com/fikrikarim/parlor

Key features

Run fully on-device voice and video AI
Stream responses sentence by sentence
Interrupt the AI using barge-in
Translate spoken input in real time
Set and track timers with audio announcements
Process background research via compatible endpoints

Use it for

Practicing spoken languages
Translating foreign speech in real time
Taking notes and transcribing speech
Managing timed tasks with voice

Alternatives to Parlor

Pipecat — Builds real-time voice and multimodal agents with frame-level control.
Xiaozhi ESP32 — It provides a hardware-based, open-source platform for building AI voice bots.
Human or AI? — Identifies AI-generated voices in real time directly inside the browser.
ClaraVerse — Protects data privacy with local execution and TEE-secured cloud inference.
Ari Helper 7 — Combines local and cloud AI models with on-device photo and film studios for one price.
Enchanted — Private, unfiltered, and multimodal chat across the Apple ecosystem.
5ire — It supports local knowledge base management and Model Context Protocol servers.
IntelliBar — Connects to multiple AI models and local models without subscription fees.
LM Studio — Runs open source language models locally with real-time voice transcription.
LibreChat — Unifies multiple AI models and tools in one customizable, open-source platform.
Chorus — Compare multiple AI chat outputs side by side without switching browser tabs.
Saanora — Chat with Mark to get answers, research, write, and code in 20+ languages.

All Parlor alternatives