Anima Workflow — V1.3

Clean, ready-to-use workflow for Anima 1.0 with smarter prompt understanding and automatic resolution handling.

Features

• Auto resolution snap — custom node rounds any input resolution to the nearest valid multiple (16/32/64), preserving aspect ratio, no seams or artifacts

• Image Saver ready — full metadata baked in, resources auto-detected on upload

• Advanced settings, better UI, Optimizer.

• Auto translate - You can write the prompt in any language; the custom node will automatically translate it.

• Easy resolution select with custom node.

Recommended settings

Setting Value Sampler res_multistep Steps 30–50 CFG 3–5 Negative worst quality, low quality, artist name, blurry, jpeg artifacts, chromatic aberration, extra fingers, missing fingers, bad hands

Required models

• anima-base-v1.0.safetensors

• qwen_3_06b_base.safetensors or qwen35_4b.safetensors (either CLIP works)

• qwen_image_vae.safetensors

• qwen3vl_4b_bf16.safetensors - For prompt enchance

Anima Workflow — V1

Clean, ready-to-use workflow for Anima 1.0 with smarter prompt understanding and automatic resolution handling.

Features

• Upgraded text encoder — swaps the stock 0.6B encoder for Qwen 3.5 4B, better at handling complex prompts

• Auto resolution snap — custom node rounds any input resolution to the nearest valid multiple (16/32/64), preserving aspect ratio, no seams or artifacts

• Image Saver ready — full metadata baked in, resources auto-detected on upload

Recommended settings

Setting Value Sampler er_sde Steps 30–50 CFG 3–5 Negative worst quality, low quality, artist name, blurry, jpeg artifacts, chromatic aberration, extra fingers, missing fingers, bad hands

Required models

• anima-base-v1.0.safetensors

• qwen_3_06b_base.safetensors or qwen35_4b.safetensors (either CLIP works)

• qwen_image_vae.safetensors

Feedback and bug reports welcome in the comments.