CIVITAI / Workflows

Anima Turbo 10-Step Text-to-Image Workflow

This workflow is for Anima Turbo 10-Step Text-to-Image. It is designed as a fast draft-generation route for testing prompts, exploring image directions, building cover concepts, and producing quick anime-style visual previews without spending too much time on every single render. The workflow uses anima_turboV10.safetensors as the main model, qwen_3_06b_base.safetensors as the text encoder, and qwen_image_vae.safetensors as the VAE. The structure is intentionally simple: positive prompt, negative prompt, latent image, KSampler, VAE Decode, and Save Image. This makes it easy to understand, easy to modify, and suitable for fast online use. The default sampling setup is 10 steps, CFG 1.0, euler sampler, simple scheduler, and full denoise. This is important: this Turbo workflow should not be treated like a high-CFG aesthetic model. It is built for low-CFG, low-step, fast-response generation. Raising CFG too high can easily make the result worse, more rigid, or less stable. The workflow is pure text-to-image only. It does not require image input, control image, pose image, mask, or reference image. You only need to write a prompt, set the size, and generate. It is especially suitable for fast anime illustration drafts, character direction tests, thumbnail concepts, safe poster compositions, and social-media visual previews. Main features: Anima Turbo 10-Step text-to-image workflow Fast draft generation anima_turboV10.safetensors model route qwen_3_06b_base.safetensors text encoder qwen_image_vae.safetensors VAE 10-step fast sampling CFG 1.0 setup euler sampler simple scheduler 1024×1024 default latent size Positive and negative prompt control Pure text-to-image workflow Suitable for prompt testing, draft generation, cover concepts, and anime-style previews Suggested workflow: Use this workflow when you need speed first. Start with a clear subject, clean composition, readable silhouette, and simple lighting direction. Avoid making the prompt too crowded during early testing. Generate multiple drafts quickly, pick the strongest direction, then move to Aesthetic 1.0 or Aesthetic 1.0b when you need a more finished image. Keep CFG around 1.0 and control the result mainly through the prompt. ⚙️ RunningHub Workflow Try the workflow online right now — no installation required. 👉 Workflow: 【Turbo RunningHub link pending】 If the results meet your exp