BustyMindy - Wan

BustyMindy for Wan 2.2 and Wan 2.1

Mindy first appeared in Flux in a "Pink" collection and has since become a Flux model and a Pony model. Now she is available for Wan videos. This character model works well for both Wan 2.1 and 2.2. The model is a little more accurate in Wan 2.1, but of course the motion and prompt adherence is much better with Wan 2.2.

Important: This model is 100% fictitious and is not based on images of any real person. She originated with images generated from still image checkpoints and has evolved from there using only AI rendering. No images (neither actual nor AI-generated) of any real person was ever involved in the making of this model.

Mindy is a blonde woman with curly hair and a tendency to wear very short tops and shiny tight bottoms. She wears lipstick and prefers high heels. While the Flux and Pony models were trained to strongly prefer pink, this Wan model has no color preference.

Wan 2.2 Usage tips

UPDATE: For the best results, please give my workflow a try. This workflow is what I use to make all my videos using my models and it produces great results in Wan 2.2 using LoRAs that were created with the Wan 2.1 model.

Wan, especially Wan 2.2, doesn't enforce the cropped tops very strongly. If you want the full authentic Mindy look, just specify that her shirt is cropped and Wan will do the rest. See sample videos for prompt guidance.

High noise LoRAs

BustyMindy: 1.00 to 2.00

Wan22-T2V-A14B-4steps-lora-rank64-Seko-V11_high_noise_model: 1.00

Low noise LoRAs

BustyMindy: 1.00

BoobPhysics, latest version: 1.50

BouncyWalk: 0.75

lightx2v_T2V_14B_cfg_step_distill_v2_lora_rank64_bf16: 1.00

CFG at 1.0, render for 6-10 steps, half the steps for high and low noise each.

You can also use a 3-part workflow instead and this will provide better natural motion at a small cost of extra render time. Use any of the sample videos above to grab my comfyui workflow to see how that works.

I like the new Seko wan22 lightning model, but only for high noise. I use the older lightx2v model for the low noise stage. The sample videos above were almost all made with this combination.

With Wan 2.1, a strength of 1.00 is plenty. For Wan 2.2, a higher strength in the high noise pipeline will give you a more busty figure, but at the cost of a little bit of motion freedom. You will get the most motion control with a strength of 1.00, but I found in testing that you can go up to 3.00 and still get very normal and realistic motion in the videos. The low noise pipeline should be kept to 1.00 or less in most cases.