Open-source anime model Anima: Turbo version flies on consumer GPUs with 8-12 steps
Time-Teaching1926 · reddit · 2026-09-26
A Reddit user shares hands-on experience with Anima, an open-source anime image model (Base and Turbo). Anima is a small-parameter DiT-architecture model (not the old UNet), friendly to consumer GPUs, with a mature ecosystem of community workflows, custom nodes, checkpoints and LoRAs.
- Turbo needs detailers and style LoRAs out of the box, but performs very well once tuned; it is extremely fast — with an INT8 quantized version you only need 8-12 steps at CFG 1.
- Base is slower but excellent, and the official Turbo LoRA works on nearly all fine-tunes.
- Compared against illustrious, NoobAI, SDXL and Chroma: Chroma remains highly versatile thanks to its massive dataset; the author also recommends lodestones/Kroma, a new Krea 2 fine-tune by Chroma's creator.
Credits go to CircleStone Labs, Comfy Org and tdrussell.
More from Multimodal
- Sophon-Realtime-v1, a full-duplex speech-to-speech model, trends on Hugging Face — adventists-ai · 2026-11-11
- Asked Fable 5.2 to 'go wild' making a game, got a Pong clone — and loved it — ariG23498 · 2026-11-11
- Western town built with Fable 5.2: ray-traced weathering ages every board — ravisparikh · 2026-11-11
- Diffusion Week 2026 releases all 10 lectures on video generation, world models and more — LucaAmb · 2026-11-11
- Running Wan 2.2 on a RTX 3080 12GB: 5B Breaks Down, 14B Works Surprisingly Well — fixxxer2606 · 2026-11-11
- Indie dev teases Z-Image-based model 'on par with latest closed-source models' — darlens13 · 2026-11-11