How to Use PixAI: The Complete Beginner’s Guide (2026)
Your complete operator's manual for PixAI — from first generation to advanced workflows. Covers prompts, models, LoRAs, editing tools, and image-to-video.
How to Use PixAI: The Complete Beginner’s Guide
PixAI gives you a complete AI art studio in one panel — text-to-image, image-to-video, LoRAs, editing tools, and more. This guide is your operator’s manual for how to use PixAI, from your first generation to advanced workflows. Whether you’re brand new or coming back after a break, you’ll find every tool here, with deep-dive links when you want to go further.
⚡ Quick Answer
How to use PixAI in 60 seconds: sign in → click Generate → type a prompt (or upload a reference image) → choose a model → click Generate. Free credits land in your account daily. Everything else in this guide — LoRAs, editing, image-to-video, advanced settings — builds on those four steps.
📑 What’s Inside
| 01 Quick Start in 60 Seconds | 05 Power Up with LoRAs |
| 02 The Generation Panel | 06 Edit & Enhance |
| 03 Master Prompt Writing | 07 Image to Video |
| 04 Choose the Right Model | 08 Beyond Generation |
01Quick Start in 60 Seconds
The fastest way to learn how to use PixAI is to actually use it. Here’s the entire flow, end to end.
STEP 1
Sign In
Open PixAI and log in. Free credits drop in daily.
STEP 2
Click Generate
From the left sidebar or the top nav bar.
STEP 3
Type a Prompt
Describe what you want. Auto Complete suggests tags as you type.
STEP 4
Hit Generate
Adjust size and batch on the right. Done!
Make a Specific Character (Like a Game Character)
Want to generate a recognizable character — say, Kafka from Honkai: Star Rail? Two paths work well:
Path 1 — Just type the name. Many popular characters are already familiar to PixAI’s models. Type “Kafka,” and Auto Complete will suggest a full prompt. Generate, and you’ll get a strong likeness.
Path 2 — Add a character LoRA. For maximum likeness and consistency, search the LoRA marketplace for a character LoRA and add it to your task. The character LoRA you can use depends on your chosen base model — switch models, and some LoRAs may become incompatible.
Switching outfits cleanly: if you want to change a character’s clothing, delete clothing words from your prompt first, then add a clothing LoRA. Mixing clothing tags with a clothing LoRA tends to create distorted results.
💡 Pro Tip
Image size and batch count both affect credit cost. Start with 512×768 and a batch of 1 while you’re learning, then scale up.
📖 For a deeper walkthrough with screenshots, see our PixAI Quick Start guide.
02The Generation Panel — Your Cockpit
The Generation Panel is the heart of PixAI. Once you understand its four zones, you understand how to use PixAI end to end. Here’s the layout, from top to bottom-right.
📝 Zone A
Prompt-Crafting Zone
The top section is where ideas take shape. The main prompt field accepts tags, phrases, or full natural-language sentences. Below it, three prompt-assistance tools take over the heavy lifting.
TOOL 1
Auto Complete
Real-time tag suggestions as you type. Massively cuts manual input.
TOOL 2
Prompt Helper
Converts natural-language sentences into optimized prompts. Multilingual.
TOOL 3
Image to Prompt
Upload any reference image to extract a full prompt set instantly.
📖 Deep dive: 3 Must-Try Prompt Tools on PixAI →
📖 Want to write better prompts from scratch? See our SDXL Prompt Writing Guide →
⚡ Boosters & Enhancement Tools
| Tool | What it does |
|---|---|
| Face Fix | Automatic facial feature refinement |
| Quality Tag | Preset quality boosts (member-only — see Membership perks) |
| HiRes Mode | High-resolution generation (one image at a time) |
| Style Mix | Combines multiple artistic styles (Waterfront models) |
| To Video | Turns text prompts into videos using PixAI’s models and LoRAs |
🖼️ Zone B
Central Display Area: Your Generated Results
This space showcases your generated images and gives you full management options for each one.
🎬 Animate
Turn static images into movie-quality videos with image-to-video.
🔍 Describe
Reverse-extract prompts from any existing image.
🎲 Variation
Create new versions from a selected image.
✏️ Edit
Precisely edit selected images with PixAI’s editing tools.
🖼️ Use as Base Image
Set as reference for new generations (img2img workflow).
💬 Flow Edit
Edit images with simple text commands — turn into a figure, swap into selfies, more.
⬇️ Download
Save as PNG.
✨ Enhance
Upscale or improve image quality.
📢 Publish
Share with the PixAI community (and earn credits when others like or use your work).
⭐ Favorite
Save to your personal collection (top-left or in your profile).
⚙️ Zone C
Right-Side Panel: Technical Controls
The right-hand panel manages the technical decisions behind every generation.
📷 Reference Image
Upload an image to guide composition or style.
🤖 Model Selection
Choose from specialized AI models for different artistic styles. We cover model selection in Part 04.
🎯 LoRA Integration
Add specialized “training packs” for specific characters, styles, poses, or outfits. We cover LoRAs in Part 05.
📐 Size Controls
Preset dimensions or custom sizes.
🔢 Batch Settings
Generate 1 or 4 images at once.
🎮 ControlNet
Use uploaded images to dictate composition and structure.
🔧 Zone D
Advanced Settings
Located below the technical controls, this is where power users dial in precision. Most beginners can skip this until they’re comfortable with everything else.
⚙️ Sampling Methods
Different algorithmic approaches to generation.
🔄 Sampling Steps
More steps = more refinement (and more credits).
🎚️ CFG Scale
Balance creative freedom vs. strict prompt adherence.
🌱 Seed Management
Control randomness and reproduce specific results.
📋 Note Function
Save and load custom setting templates so you can replay your favorite recipes.
📖 Full deep-dive: PixAI Advanced Settings — 6 Parameters to Master AI Anime Art →
03Master Prompt Writing
Once you know how to use PixAI mechanically, the next bottleneck is prompt skill. PixAI’s models respond to tag-style prompts (think: 1girl, long hair, school uniform), but Prompt Helper also accepts full sentences and converts them automatically.
Strong prompts have three layers: quality tags (masterpiece, best quality), subject description (who, where, doing what), and style modifiers (anime style, cinematic lighting). Master those and you’ve cracked it.
The Anatomy of a Strong Prompt
Here’s a real example, broken down into its parts:
FULL PROMPT:
masterpiece, best quality, ultra detailed, 1girl, long silver hair, blue eyes, school uniform, sitting by a window, golden hour lighting, soft shadows, anime style, cinematic composition
Quality tags → masterpiece, best quality, ultra detailed
Subject → 1girl, long silver hair, blue eyes, school uniform
Setting & action → sitting by a window
Lighting → golden hour lighting, soft shadows
Style → anime style, cinematic composition
Negative Prompts: The Other Half of the Equation
Negative prompts tell the model what not to draw. Common starting set: low quality, blurry, bad anatomy, extra limbs, watermark, text. They’re especially useful for cleaning up frequent generation issues like distorted hands or unintended text overlays.
Multilingual Prompting
You don’t have to write in English. Prompt Helper supports multiple languages — write in Japanese, Korean, or Chinese, and it will convert your description into optimized tags automatically. This is especially powerful for non-English speakers who want professional results without learning a tag vocabulary.
📚 Prompt Writing Library
The fundamentals — start here.
Auto Complete, Prompt Helper, Image to Prompt.
Most models default to female. Here’s how to fix that.
→ Male Character Tag Cheat Sheet
Copy-paste-ready tag library.
→ Why AI Defaults to Drawing Girls
The bias explained — and prompt fixes.
Best models for otome-style art.
For dating-sim and visual novel scenes.
Specialized prompting for BL fans.
Pairing your original characters with canon characters in a single image.
04Choose the Right Model
PixAI offers a family of AI models, each with a distinct visual character. Picking the right base model is one of the biggest factors in your output quality.
💡 How to Choose
New? Start with Tsubaki.2 — best out-of-the-box quality. Want anime/manga style? Try Haruka or Otome (Illustrious family). Want Korean-style art? Use Serin. Going for vivid color and dynamic poses? Hinata V2.
05Power Up with LoRAs
LoRAs are like specialized “training packs” you stack on top of a base model to add specific characters, art styles, poses, or outfits. They’re how you go from generic anime to your specific character, in your specific style.
Two ways to use LoRAs:
Browse and apply: open the LoRA marketplace, find a community LoRA, add it to your task. Done.
Train your own: upload 15–30 images, configure settings, hit train. PixAI handles the rest. Membership plans include free monthly LoRA trainings.
📚 LoRA Library
Beginner-friendly fundamentals.
Tune strength like a pro.
Activate LoRAs the right way.
Complete creator’s guide.
For Tsubaki, Tsubaki.2, Serin.
Save 50% with dataset reuse.
06Edit & Enhance
Generation is just step one. PixAI’s editing suite lets you refine, fix, restyle, or upscale any image — yours or someone else’s.
TOOL 1
Flow Edit
Edit images with natural-language commands. “Make her wear glasses,” “change the background to sunset” — done. Best for casual edits.
TOOL 2
PixAI Edit Pro
Advanced AI image editor for complex edits — masks, layers, fine control. Power-user mode.
TOOL 3
Inpainting & Outpainting
Mask-based AI editing — fix small areas (inpainting) or extend the canvas beyond the original frame (outpainting).
TOOL 4
Image Enhancer
Upscale, sharpen, color-correct, denoise — 6 tools plus free AI photo filters in one place.
Working with multiple reference images? Try Reference Pro for multi-image editing with natural language.
07Image to Video
PixAI’s image-to-video lets you animate any still image into a short, movie-quality clip. Generate a character, then bring it to life — pan the camera, animate movement, add motion to backgrounds.
Two entry points:
From an image: click Animate on any generated image (or upload your own).
From scratch: use the To Video button in the prompt-crafting zone for full text-to-video.
📖 Full deep-dive with model breakdown and prompt tips: PixAI Image-to-Video Tutorial →
08Beyond Generation
Once you’ve mastered the panel itself, three more areas round out everything you need to know about how to use PixAI like a pro.
CREDITS & COST
Membership & Credit Packs
Free users get daily credits. Members get more. See plans, perks, and how the credit system works.
COMMUNITY & EARN
Elite Ambassador Program
Have 1,000+ followers? Earn up to 1M credits/week creating PixAI tutorials and social posts.
NEXT-GEN AGENT
Mio.2 — AI Creative Agent
Beyond the Generation Panel: Mio.2 is PixAI’s AI agent — describe what you want, and Mio handles the prompt, model, and LoRA choices.
✨ Looking for inspiration?
Beyond standard generation
Character sheet generators, magazine covers, virtual figures, dating-sim CGs, trading cards, desktop pets — PixAI has a growing library of specialized creative tools.
Full creative-tools showcase coming soon →
Frequently Asked Questions
Quick answers to the most common questions about how to use PixAI.
Is PixAI free to use?
Yes. PixAI is free to use. New users get daily login credits automatically, can publish artwork, and access most core features without paying. Membership plans add monthly bonus credits, faster generation, image-to-video access, and higher LoRA training quotas — but the core text-to-image generator works fully on the free tier.
How do I get free credits on PixAI?
There are five free ways to earn PixAI credits without paying. Daily login gives a small bonus that resets at 00:00 UTC. Publishing original artwork earns up to 10,000 credits per day. Sharing your work to X/Twitter earns additional credits. Getting likes from the community pays out passively. Joining seasonal events and retweet campaigns can add 100,000+ bonus credits at once. Full breakdown here →
Is PixAI safe to use?
Yes. PixAI is operated by METANOMALY PTE. LTD., a registered company in Singapore. The platform enforces content moderation policies that classify content as Normal or Sensitive, with realistic sensitive content strictly prohibited. User data is handled under standard privacy practices, and users retain rights to commercial use of generated artwork in most cases (specific model licenses may vary). The mobile app is available on the Apple App Store and Google Play.
Does PixAI have a mobile app?
Yes. PixAI has a dedicated mobile app available for both iOS (App Store) and Android (Google Play). The app supports the same core text-to-image generation, daily free credits, and community features as the web version. Note that the mobile app applies stricter content moderation than the web platform, especially for realistic-style outputs.
How is PixAI different from Midjourney?
PixAI is built specifically for anime and illustration styles, while Midjourney covers a broader range of realistic and abstract aesthetics. PixAI includes a LoRA marketplace where users can apply community-trained character and style LoRAs to base models — a workflow Midjourney does not natively support. PixAI also runs on a daily-free-credit model with a web-based interface, while Midjourney requires a paid subscription and runs primarily through Discord. For anime-focused workflows, PixAI is the more specialized tool.
What’s the difference between a base model and a LoRA?
A base model (also called a checkpoint) is the large, self-contained AI that generates images on its own. A LoRA is a small add-on file you stack on top to teach the base model a specific character, style, pose, or outfit. LoRAs are typically 10-200MB versus multi-GB for base models, and require only 15-30 training images instead of thousands. You always pick a base model first; LoRAs are optional. More about LoRAs →
How do credits work, and how much does generation cost?
Credits are PixAI’s virtual currency for generation. Most standard image generations cost a few hundred credits. Larger sizes, batch counts, HiRes mode, and image-to-video cost more. You earn free credits daily through login, publishing artwork, and community engagement. You can also buy credit packs or join a Membership plan for monthly bonuses, priority generation, and exclusive perks.
Can I use my generated images commercially?
Yes — PixAI generations are usable for personal and most commercial purposes. Users retain the rights to their creations in most cases. If your output uses an existing copyrighted character or a LoRA trained on someone else’s IP, you must follow that rights holder’s terms. Specific community-uploaded models may carry their own usage restrictions, which are noted on each model’s detail page.
Can I edit images I didn’t generate myself?
Yes. Upload any image to Flow Edit, Edit Pro, or Image Enhancer. You can also use uploads as reference images, base images for img2img workflows, or as ControlNet inputs to guide composition. Uploaded images must comply with PixAI’s content policies.
Can I train my own LoRA on PixAI?
Yes. Upload 15–30 images, follow PixAI’s training flow, and you’ll get a custom LoRA in tens of minutes (DiT models like Tsubaki.2 take approximately 70 minutes). Membership plans include 3 / 5 / 10 free trainings per month for Starter / Hobbyist / higher tiers, and dataset reuse gives you a 50% discount on training fees. Full walkthrough: Train LoRA on PixAI.
⚡ Ready to Create
You now know how to use PixAI
From your first generation to advanced LoRA training, you have the full picture. Time to open the panel and bring something into the world.
