I recently started training a LoRa to the specifics of my own Lara Croft. Though this LoRa is now trained on a handful of available Lara Croft (2026/2027) Tomb Raider: Legacy of Atlantis and Catalyst.

I generated around 200 images with her "new face body", then trained a LoRa on zImage-Turbo. This is the result. Over the next weeks, I will train the further refined LoRa on a variety of checkpoints. The trigger is either classic "Lara Croft" or "laracroft26" or both combined.