40GB+ VRAM for a 3B model. VBench 85.11 beats dedicated 14B video generators. RunPod GPU costs from $2.2/session. The 'unified' model still ships as two checkpoint files.
Tested Klein 9B + 9B NSFW LoRA on M1 Max 64GB via mflux 0.17.5: 1m51s/512, 5m37s/1024 q4, 224/224 LoRA keys match, NSFW prompts uncensored, Japanese subjects work with helper tokens.
Klein 4B / 9B / Base LoRAs aren't cross-compatible — a 9B NSFW LoRA throws 'lora key not loaded' on mflux's 4B path. The variant map, what mflux runs today, and where the working hands-on test lives.
How the de-distill training adapter works for Z-Image-Turbo LoRA learning, SDXL LoRA incompatibility with Z-Image, and caption considerations specific to the Z-Image ecosystem.
Three local image generation engines (WAI-Anima, WAI-IL/SDXL, FLUX.2 Klein 4B) tied together by a thin FastAPI wrapper that takes Japanese prompts. Ollama (gemma3:12b) handles JP→EN, ComfyUI workflows are built on the fly in Python, FLUX.2 runs as an mflux subprocess, and the whole thing is reachable from an iPhone over Tailscale.
Hands-on benchmark of FLUX.2 Klein 4B on M1 Max 64GB using mflux (MLX) and iris.c (pure C + Metal). A counter to Pruna AI's H100-only tutorial — measuring how fast Apple Silicon actually gets there.
Confirmed SeeSee21/Z-Anime is a full fine-tune of Z-Image Base, then ran the AIO version on local ComfyUI on an M1 Max 64GB to verify t2i, i2i, and how NSFW prompts pass through.
A verification log for converting color anime-style AI illustrations to manga-style monochrome. AI re-generation approaches lean to either color leakage or face drift, and pure deterministic local processing looks mechanical. Frames the next directions to try: putting a grayscale-only LoRA on Anima, and using See-through for part decomposition before mechanical composition.
With v3 captions kept as-is and only the training amount pushed up to Anima's official 12,000+ step recommendation, the direction hit rate went 100% at ep150-180, crashed to 0% at ep200, then partially recovered to 67% at ep227 — a non-monotonic curve. 600-720 exposures per training image is the sweet spot; over 800 triggers catastrophic forgetting. Learning rate 2e-5, ~11 hours / $10 of RunPod training plus a sweet-spot epoch scan.
Records of rewriting captions for the 53 training images for the WAI-Anima character LoRA retrain after side ponytail direction control failed last time. Wrote position information into natural language so Qwen3 TE could pick it up, and dropped the IL-era strategy of absorbing the entire hairstyle into the single 'kanachan' trigger by promoting hairstyle to independent Danbooru tags. Includes notes on year tag necessity, the bikini/nude swapped-caption discovery, and blazer color recognition drift.
