One workspace. Full stack.
Modurian combines multi-agent AI, an in-browser IDE, live preview, and deployment — with models and routing built in. Go from prompt to a public URL without switching tools or configuring API keys.
Multi-agent modes
Build, Plan, or Ask — specialized agent behavior for coding, architecture, or Q&A without touching files.
Smart model routing
Auto picks the best included model for each message — architecture and debugging vs fast edits. No API keys to configure.
Ready on sign-in
AI models are pre-configured in the platform. Sign in with Google and start building — we handle the infrastructure.
Full IDE workspace
Chat, file tree, syntax-highlighted code viewer, live preview, and agent status — all in one screen.
Live preview
Sandboxed iframe with device modes, live reload, CSS hot reload, console, element picker, split view, a11y scan, and screenshot diff.
One-click deploy
Ship to a public URL under /spaces — shareable links that persist across restarts.
Sandbox commands
npm install, build, test — output streams to the build log with color-coded stderr.
Starter templates
Landing page, dashboard, blog, and e-commerce scaffolds to jump-start projects.
Prompt gallery
Browse expert-crafted prompts for websites, apps, slides, and docs — copy, upvote, or start a session with one click.
Prompt optimizer (Pro)
Refine your draft into a clearer agent prompt before you send — uses your selected model; 1 credit per optimization.
Reference attachments
Attach PDFs, spreadsheets, and text files to chat for context — extracted and included in the agent prompt (not project files).
Image understanding
Paste or upload images in chat. Pro routes to vision-capable models (e.g. MiniMax M3) for screenshots, mockups, and UI feedback.
Snap-to-Build (Pro)
Upload a screenshot of any website or UI — vision AI analyzes the design and rebuilds it as responsive HTML/CSS in seconds.
GitHub push
Export as ZIP or push your project to a new GitHub repo from the workspace header.
Project memory & rules
Custom rules per project, auto-learned tech stack, and context summarization for long chats.
Agent modes
Three modes for building, planning, and Q&A — switch anytime from the chat controls.
Full power — create and edit files, run npm in the sandbox, deploy to Spaces. Use for real development.
Read-only planning — architecture, file structure, and steps. No files are written until you switch to Build.
Chat only — explanations and advice without changing your project files.
Free & Pro
Sign in with Google to start free. Upgrade in Settings when you need full models and more credits.
| Free | Pro | |
|---|---|---|
| Sign-in | ||
| Models | Auto (fast tier) | Auto + manual pick, all tiers |
| Reasoning depth | Low (optimized for speed) | Low through max |
| Monthly credits | 50 on signup | 2,000 / month |
| Checkpoints & restore | — | Included |
| Skills | 4 starters (max 3 active) | Full library |
| Plugins & schedules | — | Included |
| Prompt optimizer | — | 1 credit per use |
| Reference file attach | Included | Included |
| Image uploads | Upload (vision on Pro) | Vision models + Auto routing |
Deploy with Spaces
One click deploys your project to a public URL. Files persist under /spaces/{id} and survive server restarts — share the link in demos, docs, or with your team.
- Live preview while agents edit your project
- Sandbox npm install, build, and test with streamed logs
- Export ZIP or push to GitHub from the workspace