Pigeon Surgeons, Zero LoRAs: Moss Image Model Shows Off Built-in Templates
AxonkaiLab · reddit · 2026-11-09
Reddit user AxonkaiLab shared an image generated with the Moss image model: a flock of pigeons performing surgery on a monkey in an operating room, with surprisingly solid detail and composition.
The author stresses that no LoRAs were used at all — just the model's built-in template and simple prompts with plain words, joking about misspelling "surgeons" in the first attempt. It's a直观 showcase of out-of-the-box capability for users who don't want to train or stack LoRAs.
More from Multimodal
- Seedream 5.5 consistency test: 100 outfit swaps from one base pose for 56 cents — nikola_mr64990 · 2026-11-09
- Shape My Signature turns your name into art — no prompt skills required — umesh_ai · 2026-11-09
- AI-generated Half-Life fan scene impresses Reddit with lifelike character render — seppe0815 · 2026-11-09
- AI-animated short 'Hazard' drops hazmat-suited figures into a post-apocalyptic car graveyard — Tadeo111 · 2026-11-09
- 4-step Dreamina Marketing Studio workflow swaps products into proven ad templates — socialwithaayan · 2026-11-09
- Short film 'Tough Love' made with Runway draws praise — umpherj · 2026-11-09