Meetas: fully local meeting assistant grounds every claim in transcript evidence, runs a 27B Q4 model offline
Odd-Name-1556 · reddit · 2026-09-10
The author open-sourced Meetas, a fully local meeting assistant: record/import → transcribe locally → feed your own local LLM → ask questions, analyze, extract tasks → search across meetings and documents.
The standout is grounding: analysis is structured into summaries, decisions, tasks, risks, and open questions, but every generated claim must tie back to transcript evidence; Q&A returns citations with an evidence level.
It supports Ollama or any OpenAI-compatible local server (author runs a 27B Q4 model), plus offline faster-whisper/Parakeet transcription, speaker diarization, local embeddings with hybrid search, project/document indexing, cross-meeting comparisons, and a separate chat mode that deliberately cannot access meeting data. No cloud, no account, no telemetry; Linux-only for now. GitHub: highwinglabs/meetas.
More from Infra
- d-Matrix adopts NVIDIA NVLink Fusion for rack-scale Raptor XPU inference — BenBajarin · 2026-09-11
- AI chip startup dMatrix bets on compute only, partnering with Nvidia for the rest — BenBajarin · 2026-09-11
- DeepSeek v4.1 Flash is twice as big but not twice as smart; local users better off with v4 Flash — QuixiAI · 2026-09-10
- Podcast: China's agent rules and Baseten's latest — The Cognitive Revolution · 2026-09-10
- NVIDIA open-sources BioNeMo Inference Runtime, the GPU engine behind AlphaFold DB at million-scale — AllThingsApx · 2026-09-10
- audio.cpp: pure C++ ggml engine runs TTS, ASR, voice conversion locally, no Python — solyarisoftware · 2026-09-10