User Guide & Documentation

Prompt Input Node: Write descriptions or receive AI-refined prompts. Image Reference Node: Place uploads or existing assets as style anchors. Generation Node: Set models, aspect ratios, and quality to produce images.

Drag wires from the right output dot of an input or reference card to the left input dot of a generation card. One prompt can drive multiple sizes, and one reference image can unify an entire series.

Marquee selection with Shift multi-select, grouping and ungrouping cards, one-click Auto Layout, canvas lock/unlock, and 25%–200% zoom with Fit View.

Click any finished card to open the high-res lightbox, review candidate variants from the same run, download lossless assets, and click 'Edit Based on Image' to evolve the design.

Describe your concept in plain words. The assistant automatically structures the core subject, lighting, materials, composition, and environment into high-performance prompts.

Trigger one-click actions: 'Analyze Composition & Colors', 'Style Redesign Advice', 'Extract Prompt from Image', 'Enrich Prompt Details', and 'Generate Style Variants'.

The assistant can directly add/remove nodes, wire connections, apply auto-layout, and insert refined prompts into selected cards — with full one-step undo support.

Select any input or image card on the canvas, and the AI assistant automatically uses it as conversational context to provide tailored enhancements and variation ideas.

Choose between Standard, Utility, Pro, and Utility Pro tiers — spanning cost-effective exploratory concepts to ultra-high-definition 4K commercial delivery.

Attach up to 4 reference images (from project assets, local uploads, or Ctrl+V clipboard paste) to seamlessly inherit composition, subject contours, and color palettes.

Built-in presets for standard ratios (1:1, 3:2, 4:3), social media (XHS, WeChat, TikTok, X/Twitter, Instagram), and ecommerce (Taobao, Amazon banners), with 16px-increment custom size support.

Toggle transparent background mode to output cut-out PNGs directly. Fine-tuned for ecommerce product packshots, transparent packaging, and multi-layer composites.

All jobs run asynchronously via background workers. You can freely refresh the page, switch projects, or close your browser without interrupting in-flight generations.

Before submitting any single or batch generation, the estimated credit cost is displayed next to the button based on chosen model, size, quality, and reference inputs.

Trigger batch generation across all eligible cards on the canvas. A secondary confirmation dialog clearly outlines the asset list, settings, and total credits before processing.

If upstream model timeouts, policy filters, or network errors interrupt generation, reserved credits are automatically and fully refunded to your balance.

Open the Resource Market from the sidebar to discover curated visual workflows categorized under Creation, Business, and Lifestyle, complete with node layouts and preview samples.

Click 'Import as Project' on any market card to clone the complete node graph, prompts, and reference images into your workspace as an independent, fully editable project.

Share your best visual series by clicking 'Publish to Resource Market' on the canvas page. Fill in title, design intent, and tags to submit for review and public showcase.

Like and bookmark great resources; track your draft, pending review, and published submissions directly within the 'My Submissions' panel.

Choose Creator, Studio, or Agency credit packages. Purchased credits never expire; the system automatically deducts promotional bonus credits first.

Track top-up records, generation costs, task durations, and per-job statuses in the Account Billing tab with complete audit trails.

Generate secure API keys to integrate Tanumo generation capabilities into Cursor, Claude Desktop, DSH, and custom agents via the official @tanumo/mcp-server.

Customize display name, upload user avatars (JPG/PNG/WebP), and configure email password authentication for secure access.