the best open source ai models

Hassaku XL (Illustrious) v3.1: bright, distinct anime without Danbooru clutter

Topic: Models

By Captain

Published 2026-10-01

Ikena's Hassaku XL is a bright anime model with a clear house style. Learn why it avoids Danbooru metadata tags, its CFG and negative advice, and prompt ideas for vivid characters.

Overview

Hassaku XL (

) v3.1 by Ikena aims to be an anime model with a bright and distinct style. It was trained on images with minimal disruptive elements (no floating text, logos, speech bubbles), so it produces clean compositions by default.

Two things the author stresses: Danbooru metadata and franchise tags (highres, series names) were not used in training, so skip them; and if signatures appear, add signature to the

negative prompt

.

Reference

Name	Type	What it is
Prompt	tags	Quality tags plus descriptive tags; avoid metadata and franchise tags.
Negative	short	Add signature if needed; keep it brief.
CFG	5-7	Moderate; higher values oversaturate.
Steps / sampler	25-30, Euler a
Resolution	1024	832x1216 portraits.

AD

No install required - run AI models on the cloud

BitVector Prism is the easiest way to start: pick a model, type a prompt and generate in a clean web app, with nothing to configure. BitVector is also available on Discord and on the web (SpyGlass).

Step by step

masterpiece, best quality, ultra-detailed, very aesthetic, then subject tags.
Describe appearance and outfit; skip series names.
CFG
6, Euler a, 28 steps.
Negative: lowres, bad anatomy, signature.
Hires fix 1.5x at 0.4.

Examples

Fantasy character

masterpiece, best quality, ultra-detailed, very aesthetic, depth of field, best lighting, 1girl, solo, oni horns, red skin markings, colored skin, kimono, festival night, lanterns

Winter portrait

masterpiece, best quality, 1girl, solo, fox tail, white hair, animal ears, yellow eyes, jacket, outdoors, snow, blue sky, holding coffee, smile, upper body

Creative idea: bright poster

masterpiece, best quality, very aesthetic, 1boy, solo, skateboard, rooftop, sunset, vivid colors, dynamic angle, wind

Tips

Bright, saturated palettes are the style; mention colors.
Fantasy character types (oni, fox spirit) render well from descriptive tags alone.
Weather and sky tags add atmosphere cheaply.
Illustrious
LoRAs
at 0.7.

Troubleshooting

Signatures appear

Why it happens

Training residue.

How to fix it

Add signature to the negative prompt, as the author advises.

Prompt ignored

Why it happens

Franchise or metadata tags.

How to fix it

Replace them with descriptive tags.

AD

No install required - run AI models on the cloud

PirateDiffusion is Telegram only and built for pros: thousands of models, LoRAs and workflows driven by chat commands, with unlimited generation on a fixed price plan.

Try PirateDiffusion

Questions

Can I name characters?

Franchise tags were not trained; describe the look instead.

CFG?

5-7.

Links and sources

Civitai page

Models in this guide

STABLE DIFFUSION XL 1.0

Stable Diffusion XL 1.0 Models

Written by

Captain

Related guides

CyberRealistic Pony v17: photographic portraits with Pony control

Version 17 of CyberRealistic Pony is prized for natural portrait lighting. Learn the score chain, Cyberdelia's exact settings, lens language and ideas for cinematic portraits.

iLustMix v7 Cinematic: semi-realistic Illustrious with film lighting

GZees' iLustMix v7 Cinematic blends Illustrious anatomy with a cinematic, semi-realistic finish. Learn its Euler a / 25-step settings, quality tags and ideas for moody, movie-like frames.

CyberRealistic XL v10: super clean photorealism on SDXL

Cyberdelia's CyberRealistic XL is the SDXL checkpoint for clean, sharp photoreal images. Learn the recommended DPM++ 2M SDE Karras settings, the DMD2 fast path, upscaling advice and prompt ideas.

CounterfeitXL: clean, high-contrast anime on SDXL

gsdf's CounterfeitXL carries the beloved Counterfeit look to SDXL: crisp lines, vivid color, elegant characters. Learn its tag prompting, settings and ideas for striking anime scenes.

More guides

FLUX.1 Dev: how to prompt it, best settings and creative ideas

A practical guide to FLUX.1 Dev by Black Forest Labs: natural-language prompting, guidance and step settings, LoRA stacking, text rendering and prompt ideas that play to its strengths.

Stable Diffusion XL 1.0: prompts, settings and what it still does best

How to get the most out of the official SDXL 1.0 base model: native resolutions, CFG and sampler settings, the refiner, negative prompts and ideas for styles where SDXL still shines.

Stable Diffusion 1.5: the classic model, prompted properly

Stable Diffusion 1.5 is still worth knowing: the right resolution, CFG and sampler, how to use its huge library of LoRAs and embeddings, and creative prompt ideas suited to a 512-pixel model.

FLUX.2 Dev: prompting the newest Black Forest Labs model

FLUX.2 Dev brings larger prompts, better text, stronger realism and multi-reference editing. This guide covers the turbo workflow, guidance and resolution settings, prompt structure and ideas that use its new strengths.

Qwen-Image: long prompts, perfect text and bilingual posters

Qwen-Image is the model to reach for when the words in the picture matter. Learn how to write its long descriptive prompts, render English and Chinese text accurately, set CFG and steps, and explore layout-heavy creative ideas.

SDXL-Lightning: four-step generation on any SDXL checkpoint

ByteDance's SDXL-Lightning LoRA turns a 30-step SDXL render into a 4-step one. Learn the step counts, the CFG you must use, the sampler, and how to combine it with your favorite checkpoints and style LoRAs.

© ModelTrends.ai

|

Made in Japan

|

© 2026