Trending Model:#1Inklingthinkingmachines⬇16kTrending Model:#2Ternary-Bonsai-27B-ggufprism-ml⬇432kTrending Model:#3Bonsai-27B-ggufprism-ml⬇1405kTrending Model:#4Unlimited-OCRbaidu⬇2237kTrending Model:#5GLM-5.2zai-org⬇545kTrending Model:#6Qwythos-9B-Claude-Mythos-5-1M-GGUFempero-ai⬇2133kTrending Model:#7Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-NEO-MAX-MTP-GGUFDavidAU⬇63kTrending Model:#8krea2-identity-editconradlocke⬇0kTrending Model:#9Qwen3.6-35B-A3B-Uncensored-HauhauCS-AggressiveHauhauCS⬇1998kTrending Model:#10OvisOCR2ATH-MaaS⬇17kTrending Model:#1Inklingthinkingmachines⬇16kTrending Model:#2Ternary-Bonsai-27B-ggufprism-ml⬇432kTrending Model:#3Bonsai-27B-ggufprism-ml⬇1405kTrending Model:#4Unlimited-OCRbaidu⬇2237kTrending Model:#5GLM-5.2zai-org⬇545kTrending Model:#6Qwythos-9B-Claude-Mythos-5-1M-GGUFempero-ai⬇2133kTrending Model:#7Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-NEO-MAX-MTP-GGUFDavidAU⬇63kTrending Model:#8krea2-identity-editconradlocke⬇0kTrending Model:#9Qwen3.6-35B-A3B-Uncensored-HauhauCS-AggressiveHauhauCS⬇1998kTrending Model:#10OvisOCR2ATH-MaaS⬇17k

Anima Base v1.0 Spawns Anime Art Straight from Text Prompts

By vramkickedin | | 2 min read
A floating translucent anime woman rendered in linework wearing a hoody and is holding a pencil upwards.

Anima Base v1.0 is a 2 billion parameter text-to-image model built to generate anime-style and other non-photorealistic artwork. It creates illustrations with a focus on anime concepts, characters, and styles, avoiding realistic outputs entirely. The training data includes several million anime images and about 800,000 non-anime artistic images, with no synthetic data used.

CircleStone Labs and Comfy Org developed the open-source model together for the anime art community. They trained it on curated datasets with a knowledge cut-off of September 2025 to capture contemporary visual trends. The base release offers maximum flexibility and style diversity, serving as a clean foundation for further finetuning.

Native ComfyUI support and flexible prompting

Key Features
  • 2 billion parameters for diverse anime styles.
  • Works at resolutions from 512² to 1536² pixels.
  • Native ComfyUI integration with provided workflow.
  • Accepts Danbooru tags, natural language, or mixed prompts.
  • Artist tags with "@" for precise style control.
  • Trained on millions of images, no synthetic data.
  • Base model preserves flexibility for finetuning.

Digital artists, indie game developers, and anime enthusiasts can produce custom illustrations locally without cloud dependence. Hobbyists with capable consumer GPUs keep their work private while generating high-quality concept pieces. Small studios can integrate the model into their pipelines to rapidly explore characters and visual styles.

What the developers are saying

The base model intentionally avoids photorealism and may deliver unexpected results when given very short prompts. Anima-Turbo, a faster version, is coming soon, and a comparison workflow is already available for testing against models like NetaYume. Finetuning requires a light touch—the model’s built-in visual knowledge is broad, and developers strongly recommend not training the LLM adapter.

"It is focused mainly on anime concepts, characters, and styles, but is also capable of generating a wide variety of other non-photorealistic content." — Source: Hugging Face

Project Links