Observed arrival · 2026-09-14
VRAM Wiki Maps What People Actually Do with Local LLMs
A sourced catalog of 93 community-reported local-LLM setups, connecting hardware capacity to models, tools, workflows, and results.
Field notes
The catalog organizes self-reported setups by VRAM bands, model, runtime, workflow harness, and cloud dependence rather than by benchmark score. Its examples range from Qwen 27B on a 4070 Ti Super to a four-3090 workstation used for restricted client work, with entries covering coding, transcription, document review, and homelab maintenance. The page reports 93 setups, including 32 marked as replacing the cloud, while noting that the underlying details come from community submissions.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue