Uncensored AI models for your creative work_

Abliterated Large Flash

Abliterated Large Flash is Abliteration AI's unrestricted multimodal model, based on DeepSeek V4.1 Flash and further abliterated. It has 552B backbone parameters (8B active on input, 16B on output) and a 1M-token context window. It natively processes images and text, with strong reasoning, coding, and agentic performance and no content-filtering guardrails.

Context

1,000,000 tokens

Abliterated Large V2

Abliterated Large V2 is Abliteration AI's unrestricted reasoning model, based on GLM-5.3 and further abliterated and fine-tuned. Text-only with 1M context, for harder reasoning and evaluation workloads. Always reasons; depth can be set to low, high, or max.

Context

1,000,000 tokens

Aion 3.0

Aion 3.0 is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. Multiple specialized models collaborate on each response to produce stronger narrative structure and more compelling tension and conflict. It handles mature and darker themes with nuance and supports tool calling for richer interactive fiction.

Context

128,000 tokens

Aion 3.0 Mini

Aion 3.0 Mini is a multi-model roleplaying and storytelling system from AionLabs, built on the DeepSeek family of models. Multiple specialized models collaborate on each response to produce stronger narrative structure and more compelling tension and conflict at lower cost. It handles mature and darker themes with nuance and supports tool calling for richer interactive fiction.

Context

128,000 tokens

Aion 3.5

Aion 3.5, the successor to Aion 3.0, is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models, with a longer context window. Multiple specialized models collaborate on each response to produce stronger narrative structure and more compelling tension and conflict. It handles mature and darker themes with nuance and supports tool calling for richer interactive fiction.

Context

262,144 tokens

Aion 3.5 Mini

Aion 3.5 Mini is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. Multiple specialized models collaborate on each response to produce stronger narrative structure and more compelling tension and conflict at lower cost. It handles mature and darker themes with nuance and supports tool calling for richer interactive fiction.

Context

262,144 tokens

Anime (WAI)

Anime (WAI) is WAI-Illustrious, a community anime checkpoint built on Illustrious XL, for anime illustration, including uncensored adult content. It accepts prompts of up to 1,500 characters.

Chroma

Chroma is an open-weight community text-to-image model based on FLUX.1 schnell, trained for uncensored, photorealistic generation. It accepts prompts of up to 7,500 characters.

DeepSeek V4.1 Flash 1017 Abliterated

DeepSeek V4.1 Flash 1017 Abliterated running in a Trusted Execution Environment (TEE). A text-only reasoning and coding model with refusal behavior removed and a 1M-token context window. Hardware attestation evidence is available for independent verification of enclave identity and configuration.

Context

1,000,000 tokens

DeepSeek V4.1 Flash 1017 Uncensored

DeepSeek V4.1 Flash 1017 Uncensored is an abliterated, uncensored variant of DeepSeek V4.1 Flash 1017, a multimodal Mixture-of-Experts model with 552B backbone parameters (8B active on input, 16B on output) and a 1M-token context window. It natively processes images and text, with strong reasoning, coding, and agentic performance and no content-filtering guardrails.

Context

1,000,000 tokens

Reasoning

none / low / high / max

DeepSeek V4.1 Flash Uncensored

DeepSeek V4.1 Flash Uncensored is an abliterated, uncensored variant of DeepSeek V4.1 Flash, a multimodal Mixture-of-Experts model with 552B backbone parameters (8B active on input, 16B on output) and a 1M-token context window. It natively processes images and text, with strong reasoning, coding, and agentic performance and no content-filtering guardrails.

Context

1,000,000 tokens

Reasoning

none / low / high / max

Gemma 4 26B A4B Uncensored

Gemma 4 26B A4B Uncensored running in a Trusted Execution Environment (TEE). An uncensored variant of Google's Gemma 4 MoE model with 25.2B total / 3.8B active parameters, with hardware attestation evidence available for independent verification.

Context

64,000 tokens

Gemma 4 Uncensored

Gemma 4 Uncensored is an uncensored variant of Google Gemma 4 26B, a Mixture-of-Experts model with 26B total parameters and only 4B active per token. Fine-tuned for uncensored chat without content filtering, it supports 256K context, coding, and general-purpose conversation.

Context

256,000 tokens

GLM 4.7 Flash Heretic

GLM-4.7-Flash-Heretic is an uncensored experimental variant of GLM-4.7-Flash, optimized for creative freedom and unfiltered dialogue with fast inference speed.

Context

200,000 tokens

Reasoning

none / low / medium / high

Krea 2 Turbo

Krea 2 Turbo is the speed-optimized, open-weight version of Krea's Krea 2 image model, generating high-fidelity images at 1K or 2K in seconds. It suits rapid ideation and iteration.

Longcat Distilled

Longcat Distilled is the faster, distilled version of Meituan's open-weight LongCat-Video model. This image-to-video variant animates a starting image into a 5- to 30-second clip at 720p.

Longcat Distilled

Longcat Distilled is the faster, distilled version of Meituan's open-weight LongCat-Video model. This text-to-video variant generates 5- to 30-second clips at 720p.

Longcat Full Quality

Longcat Full Quality runs Meituan's open-weight LongCat-Video model at full quality, built for longer clips. This image-to-video variant animates a starting image into a 5- to 30-second clip at 720p.

Longcat Full Quality

Longcat Full Quality runs Meituan's open-weight LongCat-Video model at full quality, built for longer clips. This text-to-video variant generates 5- to 30-second clips at 720p.

Lustify SDXL

Lustify SDXL is a community fine-tune of Stable Diffusion XL, built for uncensored, adult-oriented photorealistic images. It accepts prompts of up to 1,500 characters.

Lustify v7

Lustify v7 is version 7 of the Lustify community SDXL checkpoint, built for uncensored, adult-oriented photorealistic images. It accepts prompts of up to 1,500 characters.

Lustify v8

Lustify v8 is version 8 of the Lustify community SDXL checkpoint, built for uncensored, adult-oriented photorealistic images. It accepts prompts of up to 1,500 characters.

MiniMax H3

MiniMax H3's image-to-video variant animates a first frame, with an optional end frame, into a 5- to 15-second clip at up to 2K with native audio.

MiniMax H3

MiniMax H3, also called Hailuo 3, is MiniMax's video model. This text-to-video variant generates 5- to 15-second clips at up to 2K with native audio.

MiniMax H3 Enhanced

MiniMax H3 Enhanced is a version of MiniMax's H3 (Hailuo 3) tuned for uncensored output. This text-to-video variant generates 5- to 15-second clips at up to 2K with native audio.

MiniMax H3 Max

MiniMax H3 Max's image-to-video variant animates a first frame, with an optional end frame, into a 5- to 15-second clip at up to 1080p with native stereo audio.

MiniMax H3 Max

MiniMax H3 Max is a post-trained variant of MiniMax's H3 video model. This text-to-video variant generates 5- to 15-second clips at up to 1080p with native stereo audio.

MiniMax H3 Max Multi-Angle

MiniMax H3 Max Multi-Angle moves a virtual camera around a single image, with precise control of orbit, elevation and distance, producing 5- to 15-second clips at up to 1080p with native audio.

MiniMax H3 Max R2V

MiniMax H3 Max R2V generates 5- to 15-second clips at up to 1080p from reference images, videos and audio, cited in the prompt as @Image1, @Video1 and so on, with native audio.

MiniMax H3 Max Recast

MiniMax H3 Max Recast replaces the people in a source video with people from reference photos, preserving the original motion, camera, cuts and soundtrack, at up to 1080p.

MiniMax H3 Max Turbo

MiniMax H3 Max Turbo's image-to-video variant animates a first frame, with an optional end frame, into a 5- to 15-second clip at up to 1080p with native stereo audio.

MiniMax H3 Max Turbo

MiniMax H3 Max Turbo is a faster version of MiniMax H3 Max, generating 5- to 15-second clips at up to 1080p from text with native stereo audio.

MiniMax H3 R2V

MiniMax H3 R2V generates 5- to 15-second clips at up to 2K from reference images, videos and audio, cited in the prompt as @Image1, @Video1 and so on, with native audio.

MiniMax H3 R2V Enhanced

MiniMax H3 R2V Enhanced generates uncensored 5- to 15-second clips at up to 2K from reference images and audio, with native audio.

Ovi

Ovi is Character.AI's open-weight audio-video model, based on Wan, that animates a starting image into a 5-second clip with native audio. It runs fast and is served uncensored.

Qwen 3.6 Plus Uncensored

Qwen 3.6 Plus Uncensored is an uncensored variant of Alibaba's Qwen 3.6 Plus, a reasoning model with exceptional performance across coding, reasoning, and general knowledge tasks. Features mixed reasoning, function calling, and multimodal input support.

Context

1,000,000 tokens

Qwen 3.8 27B

Qwen 3.8 27B is a native vision-language dense model with 27B parameters. It improves coding, professional work, research, and long-horizon agentic tasks, with flexible thinking control and image and video understanding. It supports a native 262K-token context window.

Context

262,144 tokens

Reasoning

none / low / medium / xhigh

Qwen Edit Uncensored

Qwen Edit Uncensored is an uncensored image editor based on Alibaba's Qwen image editing model, with improved quality and precision. It changes a source image from a natural-language instruction, without masks.

Qwen Image 2.1 Pro

Qwen Image 2.1 Pro is the full-capacity Qwen-Image 2.1 model from Alibaba, compact and efficient while generating at 1K or 2K. It pairs with Qwen Image 2.1 Pro Edit, which takes up to 10 reference images.

Qwen Image 2.1 Pro

Qwen Image 2.1 Pro Edit is the full-capacity Qwen-Image 2.1 editor, handling local edits and compositions of people and products from up to 10 reference images at 1K or 2K.

Qwen Image 2.1 Turbo

Qwen Image 2.1 Turbo is the fast, cost-effective tier of Alibaba's open-weight Qwen-Image 2.1 series, balancing quality and speed at 1K or 2K. It pairs with Qwen Image 2.1 Turbo Edit.

Qwen Image 2.1 Turbo

Qwen Image 2.1 Turbo Edit is the fast tier of Alibaba's Qwen-Image 2.1 for editing: local edits, restyles and compositions from up to 6 input images at 1K or 2K.

Qwen Image 3

Qwen Image 3 is the standard tier of Alibaba's Qwen Image 3.0, offering cost-efficient text-to-image generation at 1K or 2K with strong text rendering. It pairs with Qwen Image 3 Edit.

Qwen Image 3

Qwen Image 3 Edit is the standard tier of Alibaba's Qwen Image 3.0 for natural-language editing, cost-efficient and able to use multiple reference images at 1K or 2K.

Qwen Image 3 Pro

Qwen Image 3 Pro is the Pro tier of Alibaba's Qwen Image 3.0, built for close prompt adherence and strong text rendering at 1K or 2K. It pairs with Qwen Image 3 Pro Edit.

Qwen Image 3 Pro

Qwen Image 3 Pro Edit is the Pro tier of Alibaba's Qwen Image 3.0 for natural-language editing with multiple reference images, at 1K or 2K.

Role Play Uncensored

Optimized for creative roleplay scenarios with maximum freedom. Designed for immersive storytelling, character interactions, and open-ended creative writing.

Context

128,000 tokens

Seedream V4.5

Seedream V4.5 is ByteDance Seed's image model, served uncensored, with strong composition and detail. It pairs with Seedream V4.5 Edit for natural-language editing.

Seedream V4.5

Seedream V4.5 Edit is ByteDance Seed's Seedream 4.5 in editing mode, known for well-composed results from natural-language instructions, with up to 6 input images.

Seedream V5 Lite

Seedream V5 Lite is the fast tier of ByteDance Seed's Seedream 5.0 image models, giving high-quality images at lower latency. It pairs with Seedream V5 Lite Edit.

Seedream V5 Lite

Seedream V5 Lite Edit is the fast tier of ByteDance Seed's Seedream 5.0 for natural-language editing, with up to 6 input images.

Seedream V5 Pro

Seedream V5 Pro is the flagship tier of ByteDance Seed's Seedream 5.0, generating high-quality images at 1K or 2K. It pairs with Seedream V5 Pro Edit.

Seedream V5 Pro

Seedream V5 Pro Edit is the flagship tier of ByteDance Seed's Seedream 5.0 for natural-language editing, at 1K or 2K with up to 6 input images.

Seedream V5.5

Seedream V5.5 is ByteDance Seed's Seedream 5.5, generating high-quality images at 1K, 1.5K, or 2K with improved realism and text rendering. It pairs with Seedream V5.5 Edit.

Seedream V5.5

Seedream V5.5 Edit is ByteDance Seed's Seedream 5.5 for natural-language editing, at 1K, 1.5K, or 2K with up to 10 input images.

Uncensored 1.2

Uncensored 1.2 is designed for maximum creative freedom and authentic interaction. Built for open-ended exploration, roleplay, and unfiltered dialogue with improved capabilities over 1.1.

Context

128,000 tokens

Wan 2.2 Enhanced

Wan 2.2 Enhanced is a version of Alibaba's Wan 2.2 for uncensored image-to-video, animating a starting image into a 5- or 8-second clip at up to 720p.

Wan 2.5 Preview

Wan 2.5 Preview is a preview release of Alibaba's Wan 2.5, served uncensored. This image-to-video variant animates a starting image into a 5- or 10-second clip at up to 1080p with native audio.

Wan 2.5 Preview

Wan 2.5 Preview is a preview release of Alibaba's Wan 2.5, served uncensored. This text-to-video variant generates 5- or 10-second clips at up to 1080p with native audio.

Wan 2.6

Wan 2.6 is Alibaba's video model with audio support. This image-to-video variant animates a starting image into a 5-, 10- or 15-second clip at up to 1080p, with optional audio, and can follow an audio track you provide.

Wan 2.6

Wan 2.6 is Alibaba's video model with audio support. This text-to-video variant generates 5-, 10- or 15-second clips at up to 1080p, with optional audio, and can follow an audio track you provide.

Wan 2.6 Flash

Wan 2.6 Flash is a fast version of Alibaba's Wan 2.6 for image-to-video, animating a starting image into a 5-, 10- or 15-second clip at up to 1080p with native audio.

Wan 2.7

Wan 2.7's image-to-video variant animates a starting image, or a first-and-last-frame pair, into an uncensored 5-, 10- or 15-second clip at up to 1080p. It can follow an audio track you provide.

Wan 2.7

Wan 2.7 is Alibaba's video model, served uncensored, for high-fidelity text-to-video in 5-, 10- or 15-second clips at up to 1080p. It can follow an audio track you provide.

Wan 2.7 Edit

Wan 2.7 Edit changes an existing video from text instructions, with an optional reference image, at up to 1080p. It's the uncensored editing variant of Alibaba's Wan 2.7.

Wan 2.7 Enhanced

Wan 2.7 Enhanced is a version of Alibaba's Wan 2.7 for uncensored video with native audio. This image-to-video variant animates a starting image into a 5-, 10- or 15-second clip at up to 1080p.

Wan 2.7 Enhanced

Wan 2.7 Enhanced is a version of Alibaba's Wan 2.7 for uncensored video with native audio. This text-to-video variant generates 5-, 10- or 15-second clips at up to 1080p.

Wan 2.7 Reference

Wan 2.7 Reference generates uncensored 5- or 10-second clips at up to 1080p from reference images, keeping characters and subjects consistent, with native audio.

Wan 3.0

Wan 3.0's image-to-video variant animates a strict first frame, or a first-and-last-frame pair, into a 2- to 30-second clip at up to 1080p, with optional native audio.

Wan 3.0

Wan 3.0 is Alibaba's video model, generating 2- to 30-second clips at up to 1080p from text, with optional native audio.

Wan 3.0 Prime

Wan 3.0 Prime's image-to-video variant quickly animates a strict first frame, or a first-and-last-frame pair, into a 2- to 30-second clip at up to 1080p.

Wan 3.0 Prime

Wan 3.0 Prime is a high-speed version of Alibaba's Wan 3.0, generating 2- to 30-second clips at up to 1080p from text much faster, with optional native audio.

Wan 3.0 Prime Pro

Wan 3.0 Prime Pro's image-to-video variant quickly animates a strict first frame, or a first-and-last-frame pair, into a 2- to 30-second clip at up to 4K.

Wan 3.0 Prime Pro

Wan 3.0 Prime Pro combines the speed of Wan 3.0 Prime with the 4K output of Wan 3.0 Pro, generating 2- to 30-second clips from text, with optional native audio.

Wan 3.0 Prime Pro Reference

Wan 3.0 Prime Pro Reference is the fastest 4K reference variant of Wan 3.0, building 2- to 30-second clips from reference images, videos and audio with consistent subjects.

Wan 3.0 Prime Reference

Wan 3.0 Prime Reference is the high-speed reference variant of Wan 3.0, building 2- to 30-second clips at up to 1080p from reference images, videos and audio with consistent subjects.

Wan 3.0 Pro

Wan 3.0 Pro's image-to-video variant animates a strict first frame, or a first-and-last-frame pair, into a 2- to 30-second clip at up to 4K.

Wan 3.0 Pro

Wan 3.0 Pro is the highest-resolution tier of Alibaba's Wan 3.0, generating 2- to 30-second clips at up to 4K from text, with optional native audio.

Wan 3.0 Pro Reference

Wan 3.0 Pro Reference builds 2- to 30-second clips at up to 4K from reference images, videos and audio, keeping subjects consistent at the pixel level.

Wan 3.0 Reference

Wan 3.0 Reference generates 2- to 30-second clips at up to 1080p from reference images, videos and audio, keeping subjects consistent at the pixel level.