Getting 2K AI art to a 10K print master: which tiled upscale workflow survives without hallucinating
Joseph1749 · reddit · 2026-09-15
An AI artist wants to take 2000×2000px generated artwork up to 10000×10000px for a 35-inch square print, without losing composition or style. Requirements: selective inpainting of small AI artifacts (e.g. missing limbs), no wholesale hallucination during upscaling, and suitability for painterly/stylized work rather than photos. The post lists community options — Tiled Diffusion, Tile ControlNet, Ultimate SD Upscale, SeedVR2, Tiled KSampler — and asks which pipeline best fits this use case.
More from coding & agent
- Solo dev argues LLMs shouldn't be the default center of an agent runtime — HmmmThisIsOdd · 2026-09-15
- How Much Code Is Throwaway Now? Engineers Rethink Project Lifespans in the AI Era — rakyll · 2026-09-15
- Translation MCP team: the hard part wasn't the server, it was getting the model to call it — Pure-Ad7786 · 2026-09-15
- Developer runs Claude inside Siri on macOS 27 via Apple's hidden model-provider support, open-sourced — Teknium · 2026-09-15
- OpenResearch tops GitHub trending, turns Claude Code and Codex into research agents — TheMoonMidas · 2026-09-15
- CoreSpeed Launches: One MCP Endpoint for Tools, Memory and Agent Budgets — Kind-Atmosphere9655 · 2026-09-15