Models by capability
Uncensored AI models
Uncensored and abliterated checkpoints refuse less often, which helps roleplay, creative writing and agent chains. Every model below uses the same endpoint, so switching is a one-line change.
43 models
Models
43 models
- DeepSeek: DeepSeek V4.1 Flash HereticFeaturedUncensoredReasoningVision512K context
Heretic checkpoint of DeepSeek V4.1 Flash 1017, served by Sable at original mixed precision (FP8 dense/Engram and MXFP4 experts). Supports text and image input, tools, structured output, and optional reasoning. Separat…
- Z.ai: Apex Flash 1 AbliteratedFeaturedUncensoredReasoningVision1M context
Apex Flash 1 Abliterated is an experimental derivative of Cantina Security and Yeta's Apex Flash 1 with modified refusal behavior.
- DeepSeek: DeepSeek V4.1 Flash UncensoredFeaturedUncensoredReasoningVision1M context
A lower-refusal finetune of DeepSeek V4.1 Flash with text and image input, optional reasoning, and tool calling. Supports a context window of up to 1M tokens.
- Xiaomi: MiMo V2.6 Flash UncensoredFeaturedUncensoredReasoningVision1M context
A lower-refusal MiMo V2.6 Flash finetune with a 1M-token context window, separate reasoning output, and tool calling. Supports image input and adjustable thinking effort.
- Z.ai: GLM 5.3 Flash UncensoredFeaturedUncensoredReasoningVision1M context
- Qwen: DeepSeek V4 Pro Qwen3.5 9B Uncensored HereticUncensored33K context
Uncensored Heretic version of Jackrong's DeepSeek V4 Pro Qwen3.5 9B, a Qwen3.5 9B model distilled from DeepSeek V4 Pro on about 250,000 math and STEM samples. It answers directly without a thinking phase. It accepts an…
- DeepSeek: DeepSeek V4.1 Flash Heretic ThinkingUncensoredReasoningVision512K context
Heretic checkpoint of DeepSeek V4.1 Flash 1017, served by Sable at original mixed precision (FP8 dense/Engram and MXFP4 experts). Supports text and image input, tools, structured output, and optional reasoning. Separat…
- DeepSeek: DeepSeek V4.1 Flash 0910 UncensoredUncensoredReasoningVision524K context
Uncensored DeepSeek V4.1 Flash with text and image input, optional reasoning, and tool calling. The date label is retained for compatibility. Served only by its own checkpoint; it may be unavailable while that host is …
- DeepSeek: DeepSeek V4.1 Flash 0910 Uncensored ThinkingUncensoredReasoningVision524K context
Uncensored DeepSeek V4.1 Flash with text and image input, optional reasoning, and tool calling. The date label is retained for compatibility. Served only by its own checkpoint; it may be unavailable while that host is …
- DeepSeek: DeepSeek V4.1 Flash 1017 UncensoredUncensoredReasoningVision1.05M context
Uncensored DeepSeek V4.1 Flash with text and image input, optional reasoning, and tool calling. The date label is retained for compatibility. Served only by its own checkpoint; it may be unavailable while that host is …
- DeepSeek: DeepSeek V4.1 Flash 1017 Uncensored ThinkingUncensoredReasoningVision1.05M context
Uncensored DeepSeek V4.1 Flash with text and image input, optional reasoning, and tool calling. The date label is retained for compatibility. Served only by its own checkpoint; it may be unavailable while that host is …
- Abliteration: Abliterated Large FlashUncensoredReasoningVision1.05M context
Abliterated Large Flash is an uncensored reasoning model based on DeepSeek V4.1-Flash-1017. It accepts text and images and returns text, with tool calling, JSON mode, JSON Schema output, and automatic prompt caching. I…
- Qwen: Qwen 3.8 27B Uncensored FastUncensoredReasoningVision1.05M context
NVFP4 build of Qwen3.8-27B by Alibaba Cloud, with refusal removal based on Huihui-Qwen3.8-27B-abliterated. Quantized and calibrated , with unpublished weights under the Model License. Supports text and image input, too…
- Google: Gemma 4 26B A4B Chimera X HereticUncensoredReasoningVision262K context
Chimera X Heretic is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.
- DeepSeek: DeepSeek V4.1 Flash Uncensored ThinkingUncensoredReasoningVision1.05M context
A lower-refusal finetune of DeepSeek V4.1 Flash with text and image input, optional reasoning, and tool calling. Supports a context window of up to 1M tokens.
- Community: CyberHeabsyUncensoredReasoning1.05M context
A low-refusal text model for authorized security research, roleplay, and creative writing. It preserves tool calling after refusal removal, supports optional reasoning, and has a 1M-token context window.
- Z.ai: GLM 5.3 UncensoredUncensoredReasoning1.05M context
An experimental uncensored version of the full GLM 5.3 reasoning model for chat, creative writing, coding, and tool use. It may show unusual behavior or inconsistent response quality.
- Xiaomi: MiMo V2.6 Flash Uncensored ThinkingUncensoredReasoningVision1.05M context
MiMo V2.6 Flash Uncensored with maximum thinking enabled. Supports a 1M-token context window and tool calling.
- Xiaomi: MiMo V2.6 Flash AbliteratedUncensoredReasoningVision1.05M context
A MiMo V2.6 Flash finetune with refusal-direction ablation, image input, a 1M-token context window, separate reasoning output, and tool calling.
- Abliteration: Abliterated Model Large V2UncensoredReasoning1M context
Abliteration.ai's default unrestricted large text reasoning model is derived from GLM-5.3 and weight-modified to reduce refusals compared with the base model. It supports native tool calling, structured output, automat…
- Abliteration: Abliterated ModelUncensoredReasoningVision262K context
Abliteration.ai's unrestricted multimodal reasoning model is weight-modified to reduce refusals and answer prompts the original model might reject. It supports text and image input, structured output, automatic prompt …
- Abliteration: Abliterated Model LargeUncensoredReasoning1M context
Abliteration.ai's unrestricted large text reasoning model is derived from GLM-5.2 and weight-modified to reduce refusals compared with the base model. It supports native tool calling, structured output, automatic promp…
- Qwen: Qwen 3.8 27B UncensoredUncensoredReasoningVision1.05M context
Qwen 3.8 27B Uncensored is an FP8 open-weight multimodal model LoRA-tuned for fewer refusals across chat, coding, reasoning, tool use, and long-context work.
- Qwen: Qwen 3.8 27B Uncensored ThinkingUncensoredReasoningVision1.05M context
Qwen 3.8 27B Uncensored with thinking enabled for more deliberate creative work, coding, multimodal analysis, tool use, and long-context problem solving.
- Google: Gemma 4 26B A4B Uncensored PrivateUncensored66K context
Gemma 4 26B A4B Uncensored Heretic is an MoE model tuned to reduce refusal behavior while preserving coding, reasoning, and function-calling strengths. Image input is unavailable on this deployment.
- Qwen: Qwen3.5 27B Blossom V6.4 DerestrictedUncensoredReasoningVision262K context
Qwen3.5 27B Blossom V6.4 Derestricted is a community creative finetune for reasoning, multimodal chat, expressive writing, and roleplay.
- Qwen: Qwen3.5 27B Claude 4.6 Opus Reasoning Distilled DerestrictedUncensoredReasoningVision262K context
Qwen3.5 27B Claude 4.6 Opus Reasoning Distilled Derestricted is a community creative finetune for reasoning, multimodal chat, expressive writing, and roleplay.
- Qwen: Qwen3.5 27B Claude 4.6 Opus Reasoning Distilled Derestricted LiteUncensoredReasoningVision262K context
Qwen3.5 27B Claude 4.6 Opus Reasoning Distilled Derestricted Lite is a lighter-tuned community finetune for responsive multimodal chat, expressive writing, and roleplay.
- Google: Gemma 4 31B Gembrain Uncensored HereticUncensoredReasoningVision262K context
Gemma 4 31B Gembrain Uncensored Heretic is a community creative finetune for reasoning, multimodal chat, expressive writing, and roleplay.
- Google: Gemma 4 31B SDFT Heretic RPUncensoredReasoningVision262K context
Gemma 4 31B SDFT Heretic RP is a community creative finetune for reasoning, multimodal chat, expressive writing, and roleplay.
