ThoxMini-tier assistant: rank-16 LoRA fine-tune of Llama-3.2-3B-Instruct by THOX on 1,565 THOX-authored examples, adapter plus bundled Q4_K_M GGUF (2.02 GB). No THOX gate result published. Llama 3.2 license.
iOS chat model: grounded ThoxChat configuration (not training) over Qwen2.5-3B-Instruct. 11.32 tok/s (measured 2026-10-31). Mobile GGUF target. Not device-qualified. Apache-2.0.
llama-3.2-3B-Q4_K_M-Korean
Professional document/image model: THOX configuration (not training) of Ministral-3-14B-Instruct-2512, all inference local. THOX evaluation pending, no figure claimed. Apache-2.0.
llmfan46/gemma-4-31B-it-uncensored-heretic-GGU with Vision
llmfan46/Qwen3.6-35B-A3B-uncensored-heretic-GGUF with Vision
llama-3-Korean-Bllossom-8B-Q4_K_M
KoboldAI/LLaMA2-13B-Tiefighter
Trained on all of TARS' lines from the movie Interstellar. Based on llama3.2-3B.
A token-efficient thinking model that rivals the CoT of much larger models with only 3 billion parameters.
Qihoo 360's first-generation reasoning model, Tiny-R1-32B-Preview, which outperforms the 70B model Deepseek-R1-Distill-Llama-70B and nearly matches the full R1 model in math.
Thai legal LLM that cites the exact law name and section (มาตรา). 30B MoE, ~3B active per token — runs in ~24 GB. Best paired with retrieval (RAG). Decision support, not legal advice.
Agent llama 3.1 8B agent
The model used is a quantized version of `Llama-3-Taiwan-8B-Instruct`. More details can be found on the https://huggingface.co/yentinglin/Llama-3-Taiwan-8B-Instruct
The model used is a quantized version of `Llama-3-Taiwan-70B-Instruct`. More details can be found on the website (https://huggingface.co/yentinglin/Llama-3-Taiwan-70B-Instruct)
The model used is a quantized version of `Llama-3-Taiwan-8B-Instruct-128k`. More details can be found on the https://huggingface.co/yentinglin/Llama-3-Taiwan-8B-Instruct-128k
ReadyArt/L3.3-The-Omega-Directive-70B-Unslop-v2.0 (Q4_K_M)
Fireball Llama 3.1 8B Code agent Q4
The model used is a quantized version of `Llama-3-Taiwan-8B-Instruct-DPO`. More details can be found on the website (https://huggingface.co/yentinglin/Llama-3-Taiwan-8B-Instruct-DPO)