/v/ - Video Games

Vidya Gaems

Index Catalog Archive Bottom Refresh
+
-
Options
Subject
Message

Max message length: 12000

files

Max file size: 32.00 MB

Total max file size: 50.00 MB

Max files: 5

Supported file types: GIF, JPG, PNG, WebM, OGG, and more

E-mail
Password

(used to delete files and posts)

Misc

Remember to follow the Rules

The backup domains are located at 8chan.se and 8chan.cc. TOR access can be found here, or you can access the TOR portal from the clearnet at Redchannit 3.0 (Temporarily Dead).

Christmas Collaboration Event
Volunteers and Ideas Needed!


8chan.moe is a hobby project with no affiliation whatsoever to the administration of any other "8chan" site, past or present.

Reminder that 8chan.se exists, and feel free to check out our friends at: Comics, Anime, Weekly Shonen Jump, /b/ but with /v/ elements, Official 8chan server: mumble.8ch.moe:64738

(5.38 MB 1280x720 Lain bored inside.mp4)

(1.63 MB 2072x1608 Super Mafia Bros.jpg)


(3.90 MB 864x1280 Selvaria's Big Bread.mp4)

(238.39 KB 1024x1536 Vintage Guts Photograph.jpg)

AI General #15: AFK Edition Anonymous 10/23/2025 (Thu) 16:45:25 Id: 8a21eb No. 1875633
Image & Video Gen, Chatbot, Voice Cloning [Copied from previous thread] ── ✦ ✦ ✦ ── Use this thread as a catch-all for AI related topics. Try to stick to vidya & help your fellow anons out. PLEASE SHARE EVERY NEW RESOURCE YOU FIND OR CREATE ── ✦ ✦ ✦ ── Resources: https://pastebin.com/rbYSd8mV Previous Thread: >>>/v/1530161 (archive.today is down) Previous threads were moved to >>>/ais/, but /v/ now has an archive. Request archival of old threads in /v/ meta >>640607. #13 >>>/ais/13775 |#12 >>>/ais/11183 │#11 >>>/ais/5650 │#10 >>>/ais/5079 │#9 >>>/ais/4560 │#8 >>>/ais/4017 │#7 >>>/ais/3494 │#6 >>>/ais/2422 │#5 >>>/ais/1776 │#4 >>>/ais/1051 │#3 >>>/ais/3 [500 posts before next thread]
>>1971178 Saar please spoiler that first image Might not be the most lewd but still...
(4.17 MB 1280x720 Alice Kittens.mp4)

Image-2-Video gen of an older Alice pic I made on the BetterDays checkpoint early this year: >>>/ais/5624 Since then, there have been new versions of that checkpoint that I should probably try: https://civarchive.com/models/623891?modelVersionId=2376051
(971.65 KB 3840x2160 Gwen Bad Kitty.jpg)

>>1974409 Trying Better Days v0.2 (creator listed shortcomings of v0.3.4.1 & recommended v0.2) https://civitai.com/models/623891?modelVersionId=2019115 Forgot how much better lighting & contrast could be with this checkpoint. Little less struggle with anatomy issues compared to NAI merges. Struggled with desks legs & wall structure a lot though & I couldn't be bothered to redraw & inpaint much more. Style / Lighting Prompt: foreshortening, depth of field, partially lit, caustics, photorealistic, chiaroscuro, dark shadows, volumetric lighting, dramatic shadows, cinematic lighting, (rimlighting, cold lighting, high contrast, dark, night, dimly lit), fog, mist, (artist:kotteri, artist:masha), Creator said use "fog" or "mist" to trigger some lighting effects, but I'm not sure that works with an interior background. Dunno how image-2-video would handle this since background is messy.
https://civitai.com/models/2163463/il-wanda-li-the-magic-school-bus?modelVersionId=2436371 Wanda Li from Magic Schoolbus. Going to try to have a LoRA ready for every day of the month till Christmas. Already got half what I need trained and a few more ready to tag.
It is tiresome trying to search AI stuff for /tg/ stuff and it always focus in D&D shit.
https://civitai.com/models/2163621/il-dorothy-ann-hudson-the-magic-schoolbus Continuing on from my Wanda LoRA from yesterday, here's one for Dorothy Ann. Tomorrow is not MSB related. >>1982258 I just wanted a few extra screenshots. It's already trained and scheduled.
Made so 2Bs. A portrait, fishing, wearing a stylish outfit and geared for combat. Tags that may be of interest: 1- anime coloring, linear hatching, cropped torso, ornate border, profile 2- anime coloring, anime screenshot, linear hatching, outline, musical note 3- traditional media, marker \(medium\), alternate costume, scarf, turtleneck shirt, waist cape, high heel boots, midriff, purse 4-5 retro, traditional media, sketch, concept art, lineart, alternate costume, seductive smile, hood down, waist cape, bullet hole
>>1979325 >>1982327 I like how well the AI changes the art style to fit the image theme. The ones holding a gun while standing against a brick wall have shading and clothing wrinkles that are not in the simpler cartoon images, and the one pic that looks like a Dragon Quest character fits that art style.
>>1982487 It's fascinating that it can (sort of) mimic the Yorha stitching pattern even on clothes that 2B has never worn.
https://civitai.com/models/2164101/il-zoe-parker-lassie-2014?modelVersionId=2437064 Zoe Parker from the 2014 Lassie cartoon. I legitimately only heard this cartoon was a thing because I saw a Pony LoRA for her. Now that it's on YouTube for free, I see I wasn't missing anything except the designs for a few girls and some wilderness background shots (and even those are gone because the second half moves to CGI shit). >>1983007 That reminds me I had a mostly done dataset for experimenting with patterns I shelved because I couldn't get Civitai to give a clear answer on if using pictures with people for non-likeness training was allowed. I should actually do that now that I finally got told I could use them.
>>1983007 Forgot to say: I note this the way its occur here is generally considered a flaw in a LoRA, where details of a particular piece of clothing get applied to anything (for example my Beebz LoRA will randomly apply her mantle's eye to anything on her torso because I didn't have enough images of her in other clothes). It just happens to come out as a nice detail on 2B though. Can probably fix it with inpainting though. >>1985133 *are gone when the second half moves to CGI shit.)
(7.96 MB 1280x720 Hermione Corrupted.mp4)

(1.39 MB 3840x2160 Hermione Corrupted.jpg)

Had issues doing image-2-video gen of Gwen Tennyson >>1976105 so I tried to make something similar with Hermione Granger using BetterDays checkpoint intending to try image-2-video again. WAN kept adding a second badge on the right side of her chest behind her knee until I added a badge on the left & flipped the image. Somehow WAN knows the Hogwarts uniform has the badge on the right side. Difficult to get all 3 video sections looking the way I want & I had to redo the last section repeatedly, but it was time to stop & just use what I had. Who knows what multiple attempts generating videos did to my electric bill. Image gen LoRAs: 0.7 weight - Hermione Granger: https://civitai.com/models/1758950 0.7 weight - torikun style: https://civitai.com/models/1009396 0.4 weight - incase new style: https://civitai.com/models/714944 Video gen LoRAs: 0.8 high 1.3 low weight - Squirt: https://civitaiarchive.com/models/1388704?modelVersionId=1572438 0.7 to 1.0 weight - Orgasm: https://civitaiarchive.com/models/1814751?modelVersionId=2373147 0.5 weight - Shaking: https://civitai.com/models/2088076/luisap-wan-22-shaking-concept 1.0 weight - Middle Finger: https://civitai.com/models/1989276 0.5 weight - Innie Pussy: https://civitaiarchive.com/models/1602000?modelVersionId=2132584 Uncertain of my video LoRA weights since I closed ComfyUI. PROTIP: Orgasm LoRA is good for expressions while Shaking LoRA helps create a more energetic orgasm. SmoothMixWAN may already know some of these concepts, but I can't be bothered to test it thoroughly. Took clips of Hermione's voice from 1st film & isolated her vocals. But she doesn't have much range in that film. Very one-note. I tried various gb (Great Britain) settings in Zonos to get her accent. It was hit n miss. Ended up using more sounds than words for the vocal gen.
Pip pip cheerio. Here's a LoRA for Florence from Thank Goodness You're Here! https://civitai.com/models/2189104?modelVersionId=2464830
>>1985866 Looks way better than the original CalArts shit, at least.
>>1986965 Too hard. Rufee cute.
https://civitai.com/models/2163783/il-phoebe-terese-the-magic-schoolbus Phoebe from Magic School Bus. Didn't mean to post her so soon after the other two, but Civitai's trainer was been a bit funny for a few days so I had to move something forward (I had left a gap I'd have filled in with something trained later, but that delay meant I've got clean datasets for other stuff but couldn't train them). >>1987673 Cute. What's Ruffeel wearing though?
>>1987680 Blue dress and mary janes.
(2.24 MB 888x1704 officialartbig.png)

>>1987751 OK. I not entirely sure if the shorts are clothing or part of her body.
(1.01 MB 960x1344 00572-522920773.png)

(914.07 KB 960x1344 00574-522920773.png)

(1006.63 KB 960x1344 00578-522920773.png)

Not video games, but these need to be shared.
>>1989208 I never asked for this
>>1989208 I just want to know if i can fuck it
>>1989208 The concept art for Impossible Creatures sure was nuts
(1.08 MB 1024x1024 00667-2977583962.png)

(1.22 MB 1024x1024 00658-2655221069.png)

(1.21 MB 1024x1024 00653-4005211477.png)

(1.07 MB 1024x1024 00652-283990126.png)

(952.65 KB 1024x1024 00645-2769930736.png)

>uploading the most vile depraved shit ever imagined Illustrious has been a godsend, I can finally images I want that don't look like dogshit. Is there anywhere I can dump a large amount of images? I have some /u/ stuff that finally looks the way I want it to, enough to actually justify a couple new tags be created in general.
(2.22 MB 416x752 dorothytowel.mp4)

(2.37 MB 416x752 dorothytowelnopan.mp4)

>>1991169 What character is this and where is she from?
>>1991575 Dorothy from Fire Emblem 6, the Binding Blade
>>1991169 What's this made with? >>1991875 Sucks she's in a game that hates bow users and even relative to other bow users isn't good.
(344.31 KB 402x662 DorothyDressing.png)

(1.93 MB 416x752 dorothytoweldress.mp4)

(1.93 MB 416x752 dorothytowelpussy.mp4)

(283.04 KB 1448x2048 yunakafattits.jpg)

(1.60 MB 464x688 yunakastripbottomless.mp4)

>>1991924 Grok's Imagine feature sometimes lets you gen lewds even though it's not supposed to. You can upload whatever image and prompt you want and sometimes it works. You only have a certain limit of generation attempts but it resets every few days
(7.00 MB 1708x960 Cirno Strong.mp4)

(1.37 MB 3840x2160 Cirno Strong Pose.jpg)

Reminder: Cirno is the strongest. Image: Dark Fantasy Oil Painting [IL LoRA]: https://civitai.com/models/1267139?modelVersionId=1469337 Video: Had WAN 2.1 version of this LoRA for a long time: https://civitai.com/models/1410898/destructive-kick-wan-i2v-14b probably should've updated it, but the old one works on WAN 2.2 regardless. Audio: Voice clone source: https://yewtu.be/watch?v=9mrEnaeOKLU
>>1992841 Oh nice, we might finally be able to make contextual Carloses.


Forms
Delete
Report
Quick Reply