PotionUI 0.0.15/0.0.16: open-source self-hosted generation with X/Y compare grids
0roborus_ · reddit · 2026-10-17
PotionUI, a free open-source self-hosted app for running image/video/audio models on your own GPU, shipped releases 0.0.15 and 0.0.16. It natively runs FLUX 3 Klein, FLUX.2, Qwen-Image 2.1, Krea-2, Z-Image, SDXL, Wan, LTX, MiniMax H3 and more, can drive ComfyUI, and works from a phone. Highlights: X/Y compare grids (pick two parameters, one Generate fills a labeled grid with per-cell timing), experimental FLUX 3 Klein with Turbo (16 steps) and Base (50 steps) checkpoints and multi-reference editing, prompt variables with independent/shared rolls, batch prompts, an in-app model browser for Hugging Face/CivitAI/OpenModelDB, 12 photo filters, auto-organize rules, and a built-in PotionAI assistant. Docker or manual install on Linux/Windows.
More from Multimodal
- PRFL ablation: full-trajectory reward feedback beats early-only or late-only training — gowthami_s · 2026-10-17
- How PRFL trains its latent reward model: noisy finished videos, no per-step labels — gowthami_s · 2026-10-17
- PRFL: reward feedback learning entirely in latent space, no VAE decoding needed — gowthami_s · 2026-10-17
- PRFL wins 67% of human evals vs RGB ReFL, though baseline may be unfair — gowthami_s · 2026-10-17
- OpenMOSS open-sources MOSS-Image: 5B DiT with CN/EN text rendering and JSON layout control — Xianbao_QIAN · 2026-10-17
- AndrewDai launches Glimpse, a 'visual thinking model' claiming frontier-level visual reasoning vs Astra — AndrewDai · 2026-10-17