Drawing Almost All of London with Claude and Seedream Took Two Weeks
Mastbubbles · reddit · 2026-08-05
A Reddit user shared their impressive project of drawing almost the entirety of London using a combination of Claude and Seedream, which took a couple of weeks to complete.
Alongside the visuals, the author provided a highly practical tip for image stitching: when generating adjacent images with Seedream, the model automatically zooms in by 20% and tilts the image by 5%. This behavior needs to be specifically accounted for during the拼接 process.
Related event: AI Used to Stitch 441 Images into London Map(2 posts)→
More from Multimodal
- Using Claude with Local ComfyUI MCP for Conversational Video Generation — EndPsychological8822 · 2026-08-06
- Is ControlNet obsolete? Krea 2-Pose showcases precise image generation control — multimodalart · 2026-08-06
- AI Video Generation Test: Recreating a Classic Sailor Moon Scene — VinceTrust · 2026-08-06
- Minimax H3 video generation stuck in 'uncanny valley', dev says — mattshumer_ · 2026-08-06
- Grok Imagine to Introduce Keyframe Support and Upgraded Image Model — chaitu · 2026-08-06
- Nano Banana Prompt Workflow: Turning Any Logo into Realistic Impasto Art — azed_ai · 2026-08-06