krea-2

514 Krea 2 Turbo prompts with generated examples and ComfyUI wildcards

Codex skill for Krea 2 prompt creation, reverse prompting, conversion, repair, evaluation, and ComfyUI/RunningHub node prompt workflows.

Prompt builder for Krea-2 in ComfyUI. Fill structured fields and get flowing prose in the format the model expects, not a keyword list.

Fork of sd-forge-negpip: experimental Krea 2 prompt weighting / NegPiP for Forge Neo (personal, AI-assisted)

Lightweight LOWVRAM optimised KREA2 webui

A simple, local Gradio workbench for checkpoint quantization, currently focused on INT8 and ConvRot conversion of .safetensors diffusion models.

🔧 ComfyUI node to offloads Krea2 transformer blocks to system RAM, without any quality loss

Community-calibrated Krea-2 LoRA training workflow (v0.4.0) for face/digital-human/character LoRAs on Ostris AI-Toolkit. TRAIN on Raw, RUN on Turbo. WebUI job takeover, gradient accumulation, 12 pitfalls documented. 训练 Krea-2 人脸/数字人 LoRA 的社区校准工作流 v0.4.0

Native MLX inference on Apple silicon: LTX-2.3 and LTX-2.5 video with audio, GMNet SDR-to-HDR still expansion, Krea 2 and FLUX.2 Klein still-image generation and editing

Local llama.cpp prompt enhancement for Krea 2 and MiniMax H3 in ComfyUI, with managed VRAM handoff.

Krea 2 Offline Image Generator and Prompt Enhancer MVP using Ollama and Streamlit.

Inference Krea-2 with ComfyUI