π 5/27/26 Update β Anima-Base v1!
The Anima v1 model (trained on Anime-Base-v1.0) is now live, optimized for blending striking anime and illustration aesthetics with the precise, structural layouts required for tabletop gaming. It is highly versatile, responding exceptionally well to both natural language narratives and classic tag-based prompts.
π Anima Prompting
Trigger:
mapcraftPrompting Style: Supports both natural language sentences and standard tag-based formatting.
Core Elements: Ensure you include structural directions like
top-down view,from above, andno humansto maintain the battle map perspective.
Natural Language Example: mapcraft. A top-down view of a mystical glowing forest clearing in a vibrant anime style. Purple flora lines the edges, with a crystal-clear pond reflecting starlight next to ancient ruins covered in glowing moss.
Tag-Based Example: mapcraft, battle map, top-down view, from above, no humans, highly detailed, anime style, mystical glowing forest clearing, vibrant purple flora, crystal-clear pond, starlight reflections, ancient ruins, glowing moss
π 7/24/25 Update β New Releases! Now available: Flux v2, SDXL, Illustrious, NoobAI EPS, and NoobAI V-Pred
πΊοΈ Mapcraft: The Ultimate TTRPG Mapmaker Generate immersive, top-down maps for fantasy, sci-fi, and modern settings. Trained on 1,000 images from references, screenshots, and past outputs, Mapcraft delivers flexible, richly detailed results.
π Flux Prompting For Flux, use the mapcraft trigger with natural language. Describe layout, setting, and atmosphere clearly from a top-down view.
Exterior Example: mapcraft. A top-down view of a medieval fort with tan tents around a stone fire pit, enclosed by a wooden palisade. Surrounded by dusty terrain, sparse shrubs, and a winding dry riverbed.
Interior Example: mapcraft. A symmetrical fantasy mansion with Arabic architecture. Domed halls, arched corridors, patterned rugs, a central courtyard with a fountain, and manicured trees outside.
π§ SDXL / NoobAI / Illustrious Prompting Trigger: mapcraft Core Tags: battle map, map, no humans, from above, top-down view, highly detailed
Exterior Maps: Add exterior, landscape, outdoors mapcraft, battle map, top-down view, from above, no humans, highly detailed, exterior, ruined desert temple, rocky cliffs, scattered bones
Interior Maps: Add interior, floorplan, indoors mapcraft, battle map, top-down view, from above, no humans, highly detailed, interior, tavern basement, cracked stone walls, hidden trapdoor
Style Tags: sci-fi, arcane, post-apocalyptic, steampunk, etc.
π§© Tip: This LoRA blends well with style-focused LoRAs. To avoid losing the top-down layout, lower other LoRA weights if needed (e.g., 0.5β0.8).
β οΈ Note: Results and prompting behavior may vary depending on the specific base models or checkpoints you pair with this LoRA. Feel free to experiment with different setups and post your results!
Description
Trained on a brand-new dataset of 1,000 images sourced from online references, in-game screenshots, and v1 outputs. Includes a trigger word for reliable activation of the LoRA.
FAQ
Comments (3)
Thank you for your hard work! One would think that it could be better represented, though... Maybe one more image would sell it better? I see a lot of creators do this thing where they include one single image of the finished product, and I just don't get it?! Batch it and do four images! Or is the problem that when you do that only one image is passable? Either way, I'm going to pass on this one.....
Working on a showcase now, got pulled away with IRL activities.
Would love it if you included a workflow! I am a noob with not a ton of experience or time.



















