Anima-Turbo-Diffusers

Quantised media packages of circlestone-labs/Anima (Anima 1.0 Turbo, v1.1 weights).

MECHA Forge

MECHA — Model Engineering & Contextual Hearth Assembly

These media packages have been forged and optimized in the MECHA Forge. Each one holds the Denoiser, Text Encoder, and VAE components packaged together for easier deployment.

Description

Anima 1.0 Turbo (v1.1 weights): distilled 2B-parameter anime/illustration text-to-image model (Cosmos-based). Fast generation with increased stability and a strong default style, at some cost to diversity. Uses a Qwen3-0.6B text encoder and the Qwen-Image VAE.

Available quantizations

Quant	Denoiser	Text encoder	VAE
FP4	FP4	FP4	FP8
FP8	FP8	FP8	FP8
INT4	INT4	INT4	INT8
INT8	INT8	INT8	INT8

The VAE is shared between formats: FP4 reuses the FP8 file, INT4 reuses the INT8 file.

Recommended settings

From the source model card (adjust to taste):

Parameter	Value
Steps	8-12
CFG / Guidance scale	1.0
Resolution	512-1536 (total pixels)
Notes	Distilled; few-step sampling; CFG 1

Example prompts

Caught in the moment (candid, slice of life):

masterpiece, best quality, score_7, safe, 1girl, solo, long hair, blue
eyes, white blouse, pleated skirt, looking at viewer, gentle smile,
smiling, city lights at night, soft bokeh, anime illustration.

Fashion photography (magazine cover style):

masterpiece, best quality, score_7, safe, 1boy, solo, short dark hair,
sharp jawline, tailored navy suit, open-collar white shirt, studio rim
lighting, bold serif typography, high contrast, cinematic, 8K, sharp
focus, GQ editorial.

Scenery:

masterpiece, best quality, score_7, safe, scenery, misty mountain ridge
at dawn, pine forests below, low clouds drifting through the valley,
first light on snow-capped peaks, anime illustration, wide angle, 8K.