INSTARAW 2.0 releases 14 ComfyUI workflows for photorealism and detector bypass
A HuggingFace release packages 14 ComfyUI workflows targeting photorealistic generation, AI detector evasion, face swap, and detail polish—recreating a setup previously sold for over $1,000.
INSTARAW 2.0 bundles 14 ComfyUI workflows built for photorealistic image generation, AI detector bypass, face swap, and final detail enhancement. Released May 16 on HuggingFace, the pack recreates a commercial workflow that had been sold for more than $1,000. The workflows rely on closed-source models—meaning users will need API access or equivalent proprietary endpoints—but the ComfyUI graph structures and node configurations are fully exposed and forkable.
The AI detector bypass workflows are the headline feature. They chain preprocessing, generation, and post-processing nodes in sequences designed to fool classifiers that flag synthetic images. Face swap and detail polish workflows round out the pack, targeting users who want to refine outputs before publication. The pack assumes familiarity with ComfyUI's node-based interface; no GUI installer or one-click setup is included.
The release arrives as demand for anti-detection workflows climbs alongside tighter content-authenticity rules on social platforms. INSTARAW 2.0's reliance on closed models limits portability—users can't swap in open weights without reworking the graphs—but the workflow logic itself is now available for study and modification. A fully open-weight version of the same pipeline would expand the pack's reach significantly. If someone ports the workflows to FLUX or SDXL fine-tunes, practitioners without proprietary API access could run the same node sequences locally, multiplying adoption among the open-model community.