Fully local voice assistant on an RTX 3060 replicates the GPT Live demo in 6.5 minutes
liampetti · reddit · 2026-09-11
A developer tested his fully local voice assistant Fulloch on the same conversation used in OpenAI's GPT Live 'Improved Intelligence' demo (flight feasibility plus food questions per destination). All on 12GB VRAM: Qwen3 1.7B ASR, Qwen3.5-9B-UD-Q4KXL with 12K context, and Pocket TTS. It finished in under 6.5 minutes; couldn't find a suitable flight route but gave solid restaurant recommendations. Source code and Docker images on GitHub (liampetti/fulloch).
More from Models
- ToneBench eval: GPT-6.1 Terra ranks #23 at writing — cheaper, not better, than GPT-5.6 Terra — OnlyProggingForFun · 2026-10-28
- Cursor user complains it silently switches to Grok, which fails basic tasks — kylegawley · 2026-10-28
- Musk says Grok Bot keeps improving, calls Vault feature a game changer — beffjezos · 2026-10-28
- Claude refuses to gather legitimately public information, sparking guardrail backlash — mkheck · 2026-10-28
- DesignArena launches Code to Video Arena: models direct 15-45s films purely in code — yshan2u · 2026-10-28
- ChatGPT's 'yes, but with an important clarification' often precedes pure nonsense — Crixusgannicus · 2026-10-28