Tested ZONOS2 on an 8GB RTX 4060 Laptop (WSL2): the 15.3GB bf16 weights run via Windows system-memory fallback, a KV-cache override, and the CUDA toolkit at ~1/20 realtime. Plus a Japanese name-accent gotcha with A/B audio.
Tested on M1 Max ComfyUI: newly free JANIMA vs Hexer Minimal Toon Anima V1 vs anima-base, one character LoRA, same seed. Hexer keeps the outfit; JANIMA adds clothes but draws the quietest backgrounds.
AFM 3 splits into 20B on-device sparse (NAND-to-DRAM weight loading) and Cloud Pro on Google Cloud NVIDIA GPU. Three Google contexts, Foundation Models API opening, and what's still unreleased.
Tested LFM2.5-1.2B-JP-202606 on M1 Max 64GB. llama.cpp Q4_K_M: 208 tok/s decode, JSON intact, model name hallucinated (LFM→FDM). Q8_0: 157 tok/s, no hallucination. Tool calls broken via GGUF.
rank128 + 20 two-character images killed the v1 ahoge bleed and body fusion on this Anima dual-character LoRA. Lap-sit stays a Qwen3 text-encoder limit; sweet spot is ep140.
One Anima (Qwen-Image DiT) LoRA, two characters, trained on RunPod: can they touch? Hugs and piggyback hold, lap-sit fuses; stacked limbs survive, interleaved break. Best at ep100, Turbo.
At Itaú, a staff engineer delivered a 4-person, 18-week project in 9 weeks with 4 AI agents — but it worked only because they knew the codebase deeply. What the case study really says about AI and team size.
Hexer Minimal Toon's new Anima V1 is a DiT checkpoint, not an SDXL one: 4.1GB bf16, separate VAE/text-encoder ComfyUI folders, Anima-only LoRA, and a non-commercial license the Civitai page doesn't show. What changes vs Illustrious v3.1 before you load it.
On Anima-Base, my character LoRA bent its legs even on standing. Adding upright references didn't fix it; cutting 36 posed full-body images did. Subtract, don't add.
