Learn Modurian
Guides and best practices for the agent workspace — models, credits, and prompts are already configured for you.
Getting started
Sign in with Google, open the workspace, pick a template or describe your app.
Included models
Pre-configured models and Auto routing — no API keys or provider accounts to set up.
Credits & plans
Free credits on signup; Pro adds monthly credits and full model control.
Modes guide
Build runs tools; Plan only reads files; Ask is chat-only with no file access.
Use cases & examples
Landing pages, dashboards, planning workflows, and deploy — with sample prompts.
Deployment (Spaces)
One-click public URLs under /spaces/{id} that persist across restarts.
Tips for better prompts
Be specific about layout, colors, and features. Use Plan mode first on large projects.
Prompt gallery
Ready-made prompts for landing pages, dashboards, e-commerce, slides, and more.
Getting started
No local API setup or provider accounts — sign in and build in the browser.
Sign in with Google
Go to Login and continue with Google. Your workspace and credits are tied to that account.
Start a project
Choose a starter template (landing, dashboard, blog, store) or describe what you want in chat.
Pick a mode
Use Plan for architecture, Build to create files and deploy, Ask for questions only.
Preview & ship
Watch the live preview update as agents work, then deploy to Spaces for a shareable URL.
Included models
Modurian hosts and configures all models — you never paste API keys or manage provider dashboards. Pick Auto to let the platform route each message, or choose a model manually on Pro.
Free uses Auto on fast-tier models with low reasoning depth for efficient exploration. Pro unlocks every model, reasoning controls, checkpoints, and higher monthly credits.
See model list and routing examplesCredits & plans
Free
50 credits
Included when you sign in with Google — enough for about 4–6 full builds or many short Ask chats (DeepSeek V4 Flash only).
Pro
2,000 credits / mo
$40/mo
All models, reasoning depth, checkpoints, and credits that refresh each billing cycle.
Credits are deducted after each agent run from weighted token usage (input + 2× output), multiplied by model cost (Flash 1×, Pro ~3×, MiniMax M3 ~7×), with minimums per mode (Build costs more than Ask). At full Pro usage, expect roughly 12,000 weighted tokens per credit on Flash — designed to align with the $40/month plan at 2,000 credits.
Deployment (Spaces)
From the workspace, deploy your project to a public URL under /spaces/{id}. Files persist on the server so links keep working after restarts. You can also export a ZIP or push to GitHub when ready.
Writing better prompts
Be specific
Include layout, colors, pages, and features — not just “make a website.” Reference styles you like when it helps.
Plan first on big work
Use Plan mode to get file structure and steps, review the plan, then switch to Build to implement.
Iterate in small steps
After the first pass, refine: “make the navbar sticky”, “add dark mode”, “fix mobile padding on the hero.”
Use templates
Templates give agents a solid base. Ask to customize branding, sections, or add pages on top.
Match mode to intent
Ask explains code without edits. Build fixes and ships. Plan never writes files — only the blueprint.
Pro: optimize before you send
Use the wand icon in the composer to tighten vague drafts into structured prompts (1 credit). Attach PDFs or images when they help.