Wan Side View Missionary - CivArchive (CivitAI Archive)

Wan Side View Missionary - Wan2.2 T2V High 1.0

NSFW

Updated for Wan 2.2. I'm pretty happy with the results of this one. The movement seems decently high speed, even when using the lightning loras. It's likely even better without the lightning loras, but I haven't checked. All of the previews were created using the lightning loras. The workflow I used for the samples can be downloaded by using the training data download.

Here's the important part of the prompt:

A man is {standing|kneeling} in front of her positioned between her legs thrusting his penis back and forth in her vagina

Description

Much better dataset and trained for wan 2.2.

FAQ

Comments (25)

duzhongzhuOct 23, 2025

CivitAI

looks good!

TwiiiistASaintTropez769Oct 23, 2025

CivitAI

Wan2.2 version looking good

DennyDan84Oct 23, 2025

CivitAI

This looks great! Thank you!

gehenon181901dfdfdOct 24, 2025· 4 reactions

CivitAI

Would have been a 10/10 if you included workflow in your samples.

dtwr434

Author

Oct 24, 2025· 2 reactions

The workflow is the training data download.

boobkake22Oct 24, 2025· 3 reactions

CivitAI

Really nice LoRA. Doesn't force white people. Face variety intact. Body type variety respected. Good motion. Penises still need some help, but that's pretty normal.

zanebeOct 24, 2025

Totally agree. Wish the author shared his wisdom and experience.

dtwr434

Author

Oct 25, 2025· 2 reactions

@zanebe There's not a ton to say, really. For this one I had a much larger variety of source videos with as much variation as possible, but clipped to the 3 or so seconds that show the movement best. I think the number of different source videos with different people is why it handles faces and races much better than other ones I've done. The limiting factor here is just how much time it takes to get the samples and caption them. I just put more time into it this time (honestly the process sucks, so it's hard to find the time/movitation to do this).

I trained the highnoise at 256 resolution on videos only, and the lownoise on the same videos, but 400 resolution, and also a bunch of images at 720 resolution.

zanebeOct 25, 2025

@dtwr434 how big dataset was? Captioned manually?

dtwr434

Author

Oct 25, 2025· 4 reactions

@zanebe I ended up with 55 3-second video clips, probably sourced from something like 35-40 videos. Sometimes I pulled 2 clips from one source video. I had 18 images included for the low noise training as well.
For captioning, I use joycaption to get the initial text created, and then I heavily edit it manually. On some images, it does better than others. Sometimes I have to replace everything except its description of the background.

zanebeOct 25, 2025

@dtwr434 Thanks!

anytimehasaikai23379Dec 26, 2025· 1 reaction

No one is forcing White people, they just want to see that.

512078862Oct 25, 2025

CivitAI

How to use this lora in I2V?

Yoorak_HuntOct 28, 2025

Just use it. You can use i2v with t2v but not the other way round.

dreamhacker69Jan 19, 2026· 1 reaction

@Yoorak_Hunt what does that mean, just use it means you cn use it but then you are saying i2v with t2v ok but not other way roung?

3417714024578Feb 26, 2026

The general idea is that T2V can be used directly in I2V, but I2V cannot be used directly in T2V. Based on my own experience, if you want to use T2V's LoRa in I2V, the T2V's LoRa weight can be adjusted slightly higher, within the range of 0.8-1. I hope this helps.

noyartOct 26, 2025

CivitAI

Possible to share what workflow you use. my t2v never look this clear. I guess my settings are wrong

dtwr434

Author

Oct 26, 2025

The training data download is the workflow.

meritrash6350Jan 27, 2026

@dtwr434 Thanks for that. When I first read it I was like "what a dick" and then checked and saw, wait, it actually IS.

1921800205q873Oct 31, 2025

CivitAI

Why are the images I generate blurry? Is it because I'm using the gguf model?

jaskenxxx119Nov 1, 2025

Try more cfg 3.0

emkayONENov 3, 2025

@jaskenxxx119 when i used the standard wan checkpoint i got this, but with secret sauce and smoothmix i didnt. Also upper the cfg helped

aadhibagavan2894541Dec 20, 2025

CivitAI

Good

wjm171732988Mar 13, 2026· 1 reaction

CivitAI

I need to recommend this lora to everyone, it comes with a penis shape, and it is different from other thick ones, it tends to be slender. It also supports all lateral postures. Not only missionaries, but also lying down, from the back, 1.2 showed good performance, 0.6 was slower. It can be linked with the lora plug-in class.

maxhilio133Mar 27, 2026· 1 reaction

CivitAI

This LORA is perfect, nice movement, intens. Charachter LORA's blend in perfectly.... great job!

LORA

Wan Video 2.2 T2V-A14B

by dtwr434

Download (Beta) View on CivitAI

concept

missionary

sex

Details

Downloads

11,005

Platform

CivitAI

Platform Status

Available

Created

10/23/2025

Updated

5/22/2026

Deleted

Files

wan2.2_t2v_highnoise_sidemissionary_v1.0.safetensors

Size:

292.59 MB

SHA256:

2e193d8a968baf4370c287876d667c37a814399030d344a6ad247014c8165704