piratediffusion posts
Posts: 0
Files: 0
Quick Video Tutorial - How to use Flux AI
In this quick video, we show you two ways to use Flux AI, the hottest new model on Graydient. You can use it on your PC or Mobile via web browser, or via Telegram. And it's really great for beginners, we cannot stress this enough. No complicated prompts or parameters required, it just understands natural language so well. In the video, we tried a ridiculous prompt and it nails it on the first try! Check out the quick video to learn how to use it. Try it today -- If you don't have an account, setup a free trial today at https://graydie...
We can't believe it's not Stable Diffusion: FLUX is available today
Introducing FLUX: Your Ultimate Creative CompanionExperience the future of AI image creation with FLUX, now available on all Graydient apps! Developed by the innovative minds at Black Forest Labs, a team of ex-Stability employees, FLUX is redefining what’s possible in the world of digital art.Why FLUX Stands Out:Versatile Mastery:From text generation and rich, vibrant colors to lifelike realism, intricate cartoons, and even those notoriously tricky hands—FLUX handles it all with ease.Beginner-Friendly:Simple to use and perfect for newcomers...
Stable Diffusion 3 is here! Try it for free
For patrons of PirateDiffusion starter, pro, and plus, we are running one server with Stable Diffusion 3 as early access. There is a special workflow command to access it, please check VIP chat boards on the MyGraydient dashboard.On the web, choose "workflows" in the Quick Create menuIn Telegram, there is a workflows command as explained in the video belowhttps://youtu.be/7WcDNYvO5XwTemplate:/workflow /run:sd3 /size:1024x1024 /seed:97809 /images:5 Outdoor night photo of a great pirate ship upon a beach, a road leading towards it, a neo...
Introducing /kimmybot and custom AI personas! Powered by SDXL and PollyGPT - You can create and customize your own
So we built an AI persona generator. Yes, and it's pure insanity. You're probably wondering why we built this, and it's because we received so many requests for it!Our community wanted to go far beyond our talking parrot, Polly. You asked for AI girlfriends, boyfriends, genius assistants, psychiatrists, and more. So we've built that. And it's here. Try our amazing demo today:To talk to Kimmy, login to your https://my.graydient.ai and click PirateDiffusion or use Telegram, and in the @piratediffusion_bot console and then ...
We've changed our name. PirateDiffusion is now Graydient
Visit our new website - https://graydient.aiThe Telegram bot is still rocking, with even more features. Check our new docs!Our WebUI is also improved and faster: The prompt stays on the screen while making images, no more reloads or page changes.Longer Prompt Weights are now supported, LCM for low guidance and low step images, and ton of new models including rapid-fire Lighting models. Over 5,000 of the most popular Checkpoints, Loras, Embeddings, and more are waiting for you. Try it today!Telegram users, reply to an image with /unified
Unlimited AI from $14/mo on our new website
We've redesigned! We now accept Paypal as well: https://graydient.aiThe PirateDiffusion Bot is still available, we just changed our name.Get 2 months free, 15% off annual plansFor a limited time, we're offering our lowest price on unlimited AI to all of like fast unlimited Stable Diffusion on all of your devices, from $14/mo prepaid available only from our new website. A bigger model selection preinstalledWe're hosting ~5,000 of the world's best AI models on-demand available across all of our applications. Whether you're cr...
Two new features! After Detailer and Showprompt Compare
After Detailer is here, and it's so easiy to use. Simply add /adetailer to your prompt and faces and hands will look better than before. That's all there is to it. The software looks at the completed render, scans for faces and hands, and applies a mini model (yolo) for each that improves the fidelity. It's similar to Facelift, but happens pre-emptively at the render runtime.Here's a little more information and an exampleThis feature works in both Stable2go webui and Telegram Bots. You can add it anywhere in your prompt and it w...
Black Friday: Get 2 months free of PirateDiffusion + Hoodie
(Click the video above to watch Episode 3 Quality Boosting tips)Our annual plan is the best value in AI - get unlimited AI images, Stable2go web app, Piratediffusion Telegram, Polly GPT Beta, and a premium black hoodie -- plus access to one of the friendliest and most fun private AI communities on earth.And if you're already a member, it's easy to upgrade to the annual plan, Patreon intelligently adds on the time without cutting your current month short.Join the latest VIP groups - many new subtopics added!New 1-Hour Video tutorial on b...
The PirateDiffusion Show, Episodes 1-3
EPISODE 3 - BOOSTING QUALITY Watch it on Youtube00:00 - We forget to remind you to win $22505:45 - Use /more to upscale and fix hands06:34 - Add strength and size in a second step17:24 - Switch art style while in mid composition23:35 - Create a reusable recipe or use a preset31:13 - Limit size in first pass to avoid duplicates42:20 - When to use guidance and strength, and why43:00 - New /vass feature for SDXL and cheat sheet48:00 - Lower concept weights to avoid glitchesEPISODE 2Watch it on YoutubeLearn how to create AI images with chat command...
New feature! One-click background removal and color replacement
By popular demand, you can now remove the background of any image, AI or from your Camera Roll, with a touch of a button. The background removal tool is now available in your WebUI tools as well as in Telegram Bot by way of the /bg command. Here is a step-by-step tutorial on how to do it with the help of blondes, muppets, and flying temples. The usual.By default, the background is set to black. However, you can pick any color, set it to full PNG transparency, and also download the mask. Advanced users can take advantage of our invert mask comma...
Stable Diffusion XL models are here!
We are proud to announce that Stable Diffusion XL 1.0 has arrived to our humble service, and that's not all. You can also now use Realistic Vision XL, DreamShaper XL, Counterfeit XL, and SDXL-based LoRAs like Loh's Cute Social Comic XL. By the end of the month we aim to have hundreds of SDXL models in the system. Can't stop, won't stop.(Gallery: XL prompts by Cauze Koopa, Mark White, Squibduck, Eggs Benadryl, & community)FAQ: Getting Started with SDXL/render your prompt <sdxl> (or another XL model from this family)SDXL is not co...
Anniversary Update! Major new features go live today
Pirate Diffusion turns oneTo be honest, we weren’t sure if anyone was going to show up when we launched our software, but we’re so happy that you all did. We’ve met so many great creators and it’s a joy to see all of your images everyday. There are too many people we want to thank to name, so let’s get straight to the drinking.We are celebrating our birthday with a massive list of new features, including a major new product release. Here's what's new:It’s not your imagination, Pirate Diffusion is now much faster.We’ve upgr...
🎁✨ We've just added & upgraded so many models
We've just added a frickin' boatload full models, loras, and inversions! These are all checkpoints, packed with 4-8 GB of training data... each! These are massive AI files with incredible range and capabilities, we can't wait for you to try these out. To see a picture of each and links to sample renders, visit our concepts browser or launch /webui from your private bot.Here's what's new:Art Universe — NEW 🔥Brick — NEW 🔥Cetus —> V4CyberReal —> V3.3Epic Realism —> V5Facebomb — NEW 🔥Fantexi — NEW 🔥Guof...
Boost your AI image quality immensely with our new Textual Inversions, and a cash contest!
We're happy to announce that powerful new Textual Inversions have arrived! We've preloaded the most popular and most powerful embeddings on the platform, so they're all ready to use right away. Follow this new guide and prompt examples to learn how to use them.Pictured: A beautiful woman with cat ears [[<bad-chilloutmix:-2>]]<chilloutmix>These negatives work similarly to very efficient, strong negative prompts, without having to type a lot of redundant crazy language like "mutated hands", and the results are just so good wit...
🎁 Major upgrades! New FaceSwap, Outpainting, better Inpaint, and more
We've just rolled out a massive batch of upgrades, here's what's new:New! Outpainting Pictured above - You can now expand the outer canvas with generative AI fill, using any inpaint model. You can also set the zoom and blur amounts.Improved InpaintingInpainting lets you replace and reprompt parts of the image by literally painting over the zone with your mouse or finger. We've made it better by: Improving upload support - inpaint images in 1 step No need to redraw your masks, you can reuse your inpaint masks The mask hides when ...
Join a Prompt Sharing Group (updated!)
One of the best reasons to join our service is meeting like-minded AI enthusiasts and getting prompt inspiration everyday, especially when you have mental fog and need help with your prompts. Here's a tour of our channels and how to find them.📯 AnnouncementsBe the first to know when new features & Models are added, get service level updates, and learn when something from our roadmap is completed.https://t.me/ai_announcements✨ InspirationBeautiful images and the full prompt. Updated semi-daily with inspirational ideas from all the group...
What's new: Weighted Prompts, LLM, big Web UI redesign, two new commands, 919(!) models, and June 15th Maintenance
Pirate Diffusion is growing! We have so many new things so let's dive right inVarious Web UI improvementsConcepts Browser update - There are now 919 models on-demand (that's 99 base + many LoRas), with improved sorting by Popularity, New, Rating, and Alphabetically. Remember to hit that adult toggle to see the full list, and within the descriptions many LoRAs contain obscure optional triggers you can now easily copy into your prompts to surface their full effects. Infinite mix/match possibilities. Tiny quality of life fix: We've als...
We've added hundreds of more LoRAs and new base checkpoint files, here's what's new
Hi, we've been busy behind the scenes uploading tons of new content for you. In fact, we've outgrown our spreadsheet as there are now 682 AI models that you can choose from and combine on the fly, right this moment. Here's some highlights of the most popular additions.Pictured above, Holokuki - An Anime model focused on high quality backgrounds and detailed waifus, with emphasis on V-Tubers. Pictured below: Beautiful Realistic Asians was updated to Version 5. We've also added MajicMix (sic) which is a digital painting model that...
Big software update! Models & LoRAs browser, and 355 models preloaded, LoRA triggers & more
Our AI models collection and browser just got bigger. Type /concepts or hit the concepts button in your private web gui to launch it.Here's what's new this week:🎁 Eleven new base models added! Try: RealFlex, Kstyle, Duchatien, DucJourney, Beautiful Real Asians, Counterfeit 3.0, MissToon, Cross Kemono, DreamShaper 5, Cetus, Meina Hentai are all excellent, mix them with your favorite LoRAs!There's now a grand total of 355 read-to-fire models on our cloud plus 7 inpaint models at your disposal. Infinite possibilities! It will take u...
Get inspired! New featured prompts channel, food pics, Web GUI update, and fresh new AI models in every style
So many good community creations, so little time. Check out the gallery of 50 top images from just last week. Truly, there's too many to post. The community is unstoppable, it's hard to believe its only been a few months and millions upon millions of images have already been created. We're blessed, this is amazing!This week, we've made it easier to find inspirational prompts by giving group owners the power to push their favorite images into a new channel called Prompt Inspiration, so you can copy and recreate these amazing imag...
Very NSFW: 18+ update - New models: URPMCUN alpha, Clarity, Liberty, Realvis 2, RevAnimated, Gay Men model, more
We're working hard to have the hottest Stable Diffusion models anywhere, and this month we've added so many more!Our models spreadsheet (now in dark mode!) is still the best place to keep track of the different styles and recommended keywords. We will be breaking each of these out into full guides in April, as we make big improvements to our websites.www.piratediffusion.com/modelsThe spreadsheet also has a new "in progress" tracker that shows what new models are being considered and tested. If you spot one that we don't have...
We have so many updates today that we have launched an update channel
First of all, thank you so much, sincerely, for your patronage. We want to do this full-time and you're helping us achieve that dream. We've been so busy building new stuff for you since our last post, and here's a big post to tell you all about it:New Channelhttps://t.me/ai_announcementsDo you ever feel like you don't know what's going on? Never again! The channel to help you manage your other 99 telegram channels has arrived. Any small improvement and bugfix is now succinctly blogged in our zero-chatter announcement channe...
New video series - learn how to create like a pro
Graydient's new Cloud Web Access companion for Pirate Diffusion Chat Bot are getting brand new features every week, so to help you keep up with the madness we have launched a Youtube Channel:https://www.youtube.com/@Graydient_AIIf you've never used our software, that's the best place to start.We're adding videos on the very basics right now, and will begin to tackle the trickier stuff in the coming days, as well as exclusive 18+ video tutorials only available to our Patreon subscribers.FREQUENTLY ASKED QUESTIONS1. Can I make rea...
(NSFW) Uber Realistic Porn Merge 1.3 is now available, and your web galleries now have passwords
We've updated our models and rolled out a new security feature!URPM 1.3 is now preloaded in your bot, right now. If its not there, here's the copy code: /concepts /copy:KX0L22XHere are the changes directly from the model'screator:- Reworked the entire recipe multiple times, blends other models in- Does leather better- Added an anime merge without losing realistic textures (danbooru tags!)- Fixed the CLIP. This will allow better prompting- Way better anatomy. Anatomy issues will still happen, but a lot less often- More porn merged in...
Over 1TB of Preloaded Models now online
Big content update this week!We've added a bunch of popular new models including Chilloutmix and Blissmix (like an Asian Hassan's Blend), DND (as in Dungeons and Dragons, a classic rpg concept) Anything v3.1 (slight 2D anime update) Babes (hourglass sexy figure illustrations) Furry (anthropomorphic anime animals) Counterfeit 2.5 (a very cute anime alternative to AV3) and Pastel from the makers of AV4, which looks like expressionistic Japanese illustrations. These are all in the beta channel, meaning they may have bugs and won't rend...
We've upgraded our AI upscaler, now 4X sharper
We're pushing a lot more pixels now when you AI upscale your images, and it also works with your regular non-AI photos, too. It has three (updated!) modes, and here's how to use them:/faceliftUse this for traditional face touch-up, similar to your smartphone's "beauty" mode/facelift /animeTo boost the pixels and sharpen fantasy art and anime, or semi-realistic images. Works great on oil paints, too.Check the attachments to compare the difference./facelift /photoAnd "photo" mode skips face retouching, so you keep faci...
FAQ: How do I change the sizes of my photos?
Available now for everyone - you can quickly change the proportions of an image with these handy shortcuts. Examples on our blogLearning how to change the aspect ratio of an image with AI is very easy, but unlike cameras and design software, there’s a few more considerations.AI models “learn” using square 1:1 aspect ratio images, so when we ask for a picture back that isn’t a square, this can result in glitches like multiple heads on a person, the same person twice in an image, or a crash: causing images to render poorly or not at all. ...
FAQ: How do I join groups?
We're all about bringing like-minded AI enthusiasts together, so we've created additional channels to share your creations. This post used to have a list of all the rooms, but there's so many of them now that it's just better to show you an example.On the left side is an icon representing each room. For example, the Test / Beginner's Room has 95 members learning Stable Diffusion for the first time. So if you're new to AI, come chat here. It's a friendly place and we have wild pandas!More ways to express yourself in P...
More GPUs online! Two new guides! How to upscale your old prompts, and exploring Dark Voyage AI model
MORE POWER TO YOUThanks for bearing with us as we grow. We want to keep the service as fast as the day you arrived, so we've brought on more servers to bring back top speeds. You should notice an improvement next time you render.LESS CASE SENSITIVEYou don't have to worry about /Render vs. /render anymore, we made the system accept both. This was driving people with Samsung keyboards crazy, so we fixed it. If you have any pet peeves like that please don't hesitate to tell us, often times we just don't know about it or it's al...
Three (3!) new high quality models added! ProtoNova, Ramna, Niji
New week, new features! We've updated our website with more tutorials and examples, check it outNew Fantasy Model: <protonova>Try the 8.0 release of Protogen Nova in early access, which we liked a lot more than 8.6 Infinity. It's a bit tricky to work with as there are multiple trigger words for different styles, so note that the trigger words go last in your prompt. You can supposedly repeat the word Protogen as well for more fantasy effects, though in our testing that seems hit or miss. Either way, it's beautiful.New models for Ani...
