HN Showcase
Back to browse
GitHub Repository

CPU-only voice agent approximating Thinking Machines' Interaction Models demo

17 starsPython

Cheap-IM – CPU-only voice agent approximating Thinking Machines' demo

by mrkn1·May 17, 2026·4 points·0 comments
Visit ProjectView on HN

AI Analysis

●●●BangerBig BrainWizardryDark Horse

Runs real-time vision-keyed voice agents on a laptop CPU without custom silicon or training.

Strengths
  • •Orchestrates YOLO11 pose detection and Silero VAD in a single async loop for sub-second reactions.
  • •Achieves 'slouch detection' and friend identification using purely off-the-shelf local models.
  • •Handles mid-conversation interrupts while background workers generate charts via API calls.
Weaknesses
  • •Still relies on remote LLM APIs (DeepInfra) rather than fully local inference for reasoning.
  • •Zero stars and no releases yet, suggesting early alpha status despite the impressive demo.
Category
AI/ML
Target Audience

Developers building local-first AI agents, hobbyists

Similar To

Thinking Machines Interaction Models · Open Interpreter · Home Assistant

Similar Projects

AI/ML●●Solid

Local Voice Assistant

This repo bundles a complete local audio loop — client captures audio, backend transcribes with Parakeet, queries a quantized Mistral LLM via Ollama, then renders speech with Kokoro or Qwen3-TTS for cloning — and reports ~1s round-trip on an RTX5070. It’s a practical, take-it-home demo for running privacy-first voice agents, though it’s still a demo: requires specific tooling (Ollama, GPU headroom), has obvious TODOs (VAD, better warmup for cloning), and isn’t reinventing the architecture.

WizardryNiche Gem
armcat
204mo ago