Guide: best uncensored local AI models per category and a 4-step Ollama setup for $0
udmrzn · x · 2026-11-10
A practical roundup of the best uncensored local models: Qwen3.8-27B Abliterated (overall), Gemma 4 12B/26B-A4B Abliterated (small GPU/fastest), Hermes 4.3-36B (no surgery), FLUX.1-dev (image), Wan 2.2 (video), GLM-5.1 Abliterated (ceiling, 754B).
Zero-cost setup in 4 steps (abliterated Gemma 4 example):
- Install a recent Ollama build (older ones lack Anthropic API and Gemma 4 support)
- ollama pull igorls/gemma-4-12B-it-heretic-GGUF — avoid random "uncensored" tags; the Heretic-abliterated pick cut refusals from 99/100 to 0/100
- ollama run <model> + /set nothink
- No API key, account, or card needed; can be plugged into Claude Code
Also linked: huihui-ai's Hugging Face page, a team specializing in model abliteration, publishing abliterated GGUF quants of GLM, Qwen, Kolibri and more.
More from Infra
- hf-mem 1.0.0 nears stable release with vLLM/SGLang-aligned memory estimates — anton_lozhkov · 2026-11-10
- Documentary visits Delta, Utah: a 4,000-person farming town hosting the world's biggest data center campus — zealcaiden · 2026-11-10
- Meta's Muse Ember matches Claude Haiku 5.5 on Intelligence Index at a third of the cost — ArtificialAnlys · 2026-11-10
- Hot agent harnesses share one thing: they run on Durable Objects — irvinebroque · 2026-11-10
- SurrealDB Earth launches serverless database on Cloudflare: 3M free requests, 330+ regions — mattturck · 2026-11-10
- B300 GPU spot rental price doubles to $4.50/hr in one day — kevinsxu · 2026-11-10