saidutta69/Qwen3-8B-heretic

Text Generation • 8B • Updated 14 days ago • 2.45k • 3

Note Best for: local agents, roleplay, and studying refusal mechanics with Qwen3's thinking/non-thinking dual mode.

saidutta69/gemma-3-12b-it-heretic

Text Generation • 12B • Updated 14 days ago • 1.77k

Note Best for: multimodal uncensored chat (text + vision) on 16-24 GB GPUs.

saidutta69/gemma-3-4b-it-heretic

Text Generation • 4B • Updated 14 days ago • 994

Note Best for: small multimodal uncensored chat on consumer GPUs and edge devices.

saidutta69/Mistral-7B-Instruct-v0.3-heretic

Text Generation • 7B • Updated 14 days ago • 2.33k • 2

Note Best for: general-purpose uncensored chat, agents, roleplay, and function-calling on consumer hardware.

saidutta69/Qwen2.5-Coder-7B-Instruct-heretic

Text Generation • 8B • Updated 14 days ago • 3.07k

Note Best for: uncensored coding agents, pair-programming, and repo-level assistance.

saidutta69/Qwen2.5-14B-Instruct-heretic

Text Generation • 15B • Updated 14 days ago • 1.49k

Note Best for: strong uncensored reasoning and instruction-following on 16-24 GB GPUs.

saidutta69/Qwen2.5-7B-Instruct-heretic

Text Generation • 8B • Updated 14 days ago • 1.07k

Note Best for: mid-size uncensored local chat, agents, and roleplay.

saidutta69/Qwen2.5-Coder-0.5B-Instruct-heretic

Text Generation • 0.5B • Updated 17 days ago • 1.22k

Note Best for: tiny uncensored code copilot on CPUs and edge devices.

saidutta69/MiniCPM5-1B-Claude-Opus-Fable5-V2-Thinking-heretic

Text Generation • 1B • Updated 14 days ago • 8.01k • 9

Note Best for: tiny uncensored Thinking model with strong tool-calling and coding.

saidutta69/MiniCPM5-1B-heretic

Text Generation • 1B • Updated 14 days ago • 2.82k • 3

Note Best for: on-device agentic tool use, code generation, and 128K long-context.

saidutta69/Qwen2.5-1.5B-Instruct-heretic

Text Generation • 2B • Updated 16 days ago • 1.42k

Note Best for: small uncensored chat on CPU or low-VRAM devices.

saidutta69/Qwen2.5-0.5B-Instruct-heretic

Text Generation • 0.5B • Updated 14 days ago • 3.47k

Note Best for: the smallest heretic — CPU-only and edge/embedded inference.

saidutta69/Llama-3.2-1B-Instruct-heretic

Text Generation • 1B • Updated 16 days ago • 1.39k

Note Best for: uncensored on-device, mobile, and edge deployment.

saidutta69/Qwen3-0.6B-heretic

Text Generation • 0.6B • Updated 17 days ago • 1.38k • 1

Note Best for: smallest Qwen3 heretic — CPU-only, edge, and refusal-mechanics testbed.

saidutta69/Qwen2.5-3B-Instruct-heretic

Text Generation • 3B • Updated 14 days ago • 1.86k • 1

Note Best for: small, fast uncensored local chat and roleplay.

saidutta69/Qwen2.5-Coder-3B-Instruct-heretic

Text Generation • 3B • Updated 14 days ago • 2.62k • 1

Note Best for: uncensored coding assistance on small machines.

saidutta69/DeepSeek-R1-Distill-Qwen-7B-heretic

Text Generation • 8B • Updated about 1 month ago • 7.13k

Note Best for: uncensored R1-style chain-of-thought reasoning on 8-12 GB GPUs.

saidutta69/DeepSeek-R1-Distill-Llama-8B-heretic

Text Generation • 8B • Updated about 1 month ago • 5.65k

Note Best for: uncensored deep reasoning on consumer hardware.

saidutta69/Meta-Llama-3.1-8B-Instruct-heretic

Text Generation • 8B • Updated about 1 month ago • 678

Note Best for: the most widely used open LLM, now without refusals.

saidutta69/Qwen3.5-9B-heretic

Text Generation • 9B • Updated about 11 hours ago • 2.42k

Note Best for: latest Qwen3.5 multilingual reasoning and agentic use on 12 GB GPUs.

saidutta69/DeepSeek-R1-Distill-Qwen-14B-heretic

Text Generation • 15B • Updated about 1 month ago • 4.98k • 2

Note Best for: largest Qwen-distilled R1 reasoning without refusals.

saidutta69/SmolLM3-3B-heretic

Text Generation • 3B • Updated 14 days ago • 1.82k

Note Best for: tiny 3B uncensored chat for edge and on-device inference.

saidutta69/Llama-3.2-3B-Instruct-heretic

Text Generation • 3B • Updated 16 days ago • 4.08k • 5

Note Best for: mid-size uncensored Llama-3.2 chat on consumer hardware.

saidutta69/Qwen2.5-7B-Instruct-1M-heretic

Text Generation • 8B • Updated 14 days ago • 3.45k • 1

Note Best for: massive 1M-token context for long-document and codebase analysis.

saidutta69/c4ai-command-r7b-12-2024-heretic

Text Generation • 8B • Updated 14 days ago • 1.94k • 3

Note Best for: multilingual RAG, retrieval-augmented generation, and tool use.

saidutta69/granite-4.1-8b-heretic

Text Generation • 8B • Updated 14 days ago • 1.39k • 2

Note Best for: enterprise-grade coding and reasoning without censorship.

saidutta69/Yi-1.5-9B-Chat-16K-heretic

Text Generation • 9B • Updated 14 days ago • 753

Note Best for: 16K-context multilingual uncensored conversation.

saidutta69/Mistral-Nemo-Instruct-heretic

Text Generation • 12B • Updated 14 days ago • 2.86k • • 4

Note Best for: 12B multilingual reasoning and long-context applications.

saidutta69/Qwen3.5-0.8B-heretic

Text Generation • 0.9B • Updated 16 days ago • 1.19k

Note Best for: tiny fast hybrid-attention Qwen3.5 on CPU and edge.

saidutta69/gemma-3-1b-it-heretic

Text Generation • 1.0B • Updated about 1 month ago • 521

Note Best for: Raspberry Pi-class and tiny edge uncensored chat.

saidutta69/DeepSeek-R1-Distill-Qwen-1.5B-heretic

Text Generation • 2B • Updated 16 days ago • 1.55k

Note Best for: R1-style reasoning at ~1 GB quantized on CPU.

saidutta69/Qwen3.5-2B-heretic

Text Generation • 2B • Updated 16 days ago • 1.33k

Note Best for: fast hybrid-attention Qwen3.5 on CPU and edge devices.

saidutta69/RaceBench-MiniCPM5-heretic

Text Generation • 1B • Updated 16 days ago • 1.9k • 1

Note Best for: 1B racebench reasoning in a hybrid Think/No-Think package for on-device and edge deployment.

saidutta69/Qwen3-VL-8B-Instruct-heretic

Image-Text-to-Text • 9B • Updated 29 days ago • 605

Note Best for: uncensored vision-language chat (text + vision) - OCR, grounding, GUI agents - on 16-24 GB GPUs.

saidutta69/Qwen2.5-VL-7B-Instruct-heretic

Image-Text-to-Text • 8B • Updated 29 days ago • 936

Note Best for: multimodal uncensored chat (text + vision) on 12-16 GB GPUs.

saidutta69/VibeThinker-3B-heretic

Text Generation • 3B • Updated 14 days ago • 1.98k

Note Best for: small uncensored Qwen2.5-Coder reasoning with VibeThinker thinking mode on CPU and low-VRAM.

saidutta69/lfm2.5-2.6b-fable5-coding-agent-heretic

Text Generation • 3B • Updated 14 days ago • 9.02k • 7

Note Best for: fast hybrid conv+attention coding agent on CPU and edge — 2.6B with LiquidAI LFM2 architecture.

saidutta69/SmolLM2-135M-Instruct-heretic

Text Generation • 0.1B • Updated 14 days ago • 974

Note Best for: the smallest heretic yet — 135M for CPU-only, edge/embedded, and on-device experiments.