Documentation
Back to homeLearn to build with Ntekuma
Ntekuma is a workspace for generating websites, documents, and media from prompts and uploads. This guide explains how to get great results, how builds work, and how to troubleshoot common issues.
Quick Start
Create a project
Click 'Create Project' and give it a name
Describe your vision
Write what you want to build in natural language
Select a model
Choose based on speed, quality, and budget
Run the build
Click build and watch your creation come to life
Iterate & refine
Request changes to polish your project
What You Can Generate
Web Pages
Static HTML/CSS/JS sites, landing pages, portfolios
UI Prototypes
Component-driven layouts and design systems
Documents
Reports, guides, checklists, and summaries
Media
Video generation with Sora-2, character consistency, TTS narration
Mobile Apps
Expo project scaffolds with setup instructions
Dashboards
Data visualization and analytics interfaces
Design-to-Code
Upload screenshots/mockups, get production React/Vue/HTML code
Auto Tests
AI-generated unit, integration, and E2E tests
Prompting Tips
Good Prompts
- ✓"Create a landing page for a coffee shop with a hero image, menu section, location map, and contact form. Use warm brown colors and modern typography."
- ✓"Build a dashboard showing sales metrics with a sidebar navigation, 4 KPI cards at the top, a line chart for revenue trends, and a data table below."
Avoid These
- ✗"Make me a website" (too vague)
- ✗"Build an app like Instagram" (too complex for one prompt)
Pro Tips
- Be explicit about the target: web page, document, video, or app scaffold
- List required sections or screens in order of importance
- Specify fonts, colors, and layout constraints when relevant
- For edits, reference the exact element and what should stay the same
- Upload reference images to show the style you want
Iterative Editing
Describe the delta, not the whole thing.
When requesting changes, tell the AI what to keep and what to modify. This helps refine instead of rebuild.
Example: "Keep the hero layout, change the CTA button to red, and replace the feature icons with illustrations."
Uploads & References
You can upload images, PDFs, or text files to guide the AI. Be specific about how each upload should be used:
Supported Formats
- • Images: PNG, JPG, WebP, SVG
- • Documents: PDF, TXT, MD
- • Data: JSON, CSV
Usage Examples
- • "Use this photo as the hero background"
- • "Follow this layout exactly"
- • "Extract the color palette from this image"
Choosing Models
| Use Case | Recommended Model |
|---|---|
| Quick drafts & iterations | GPT-5.2, Claude Sonnet 4.5 |
| Complex UI, high quality | GPT-5.2 Pro, Claude Opus 4.5 |
| Video generation | Sora-2 |
| Final polish | Claude Opus 4.5 |
Tip: Use faster models for drafts, switch to stronger models for final polish.
Platform Features
Version Control & History
Every change is tracked. Branch your project, compare versions, view diffs, and rollback to any previous state with a single click.
Real-Time Collaboration
Work together in real-time with cursor presence, synchronized editing, and chat. Perfect for team projects and pair programming.
SaaS Starter Templates
Start from 12+ production-ready templates: SaaS Starter, AI Chatbot, E-commerce, Marketplace, CRM, LMS, and more.
GitHub Two-Way Sync
Connect to GitHub repositories. Push changes, pull updates, and keep your code in sync automatically.
Custom Domains with SSL
Deploy your projects to custom domains with automatic SSL certificate provisioning and DNS verification.
AI Agent Mode
Describe a goal and let the AI autonomously plan and execute multi-step tasks: read files, write code, run commands, and iterate until done.
Downloads & Deployment
After a successful build, you can download your generated files:
- ZIP Export — Download all project files as a ZIP archive
- Individual Files — Download specific files from the file browser
- Video Export — Download generated videos in MP4 format
- Deploy to Hosting — One-click deploy to Vercel or Netlify (coming soon)
Cinema Engine
Generate cinematic videos up to 10 minutes with AI. Write a screenplay, define characters, and produce multi-scene movies with consistent characters, voice acting, soundtracks, and professional camera work.
How It Works
- Write your script — Import or write a screenplay with character descriptions, scene headings, and dialogue
- Generate storyboards — AI creates preview images for each shot (~1 credit each). Review and approve before video
- Generate video — Approved storyboards become reference frames for video generation (8 credits per shot)
- Automatic assembly — Shots are stitched into scenes, episodes, and the final output with FFmpeg
AI Camera Tool
Professional camera simulation is built into every shot. The engine auto-selects optimal settings, or you can specify them in your script.
Script Camera Directives
Add camera directives directly in your script. The parser auto-detects them:
If no camera directives are specified, the AI Camera Tool auto-selects optimal settings based on shot type, mood, and style.
Academic Source Search
Ntekuma automatically searches 250M+ academic papers across OpenAlex, Semantic Scholar, and CrossRef to ground your text in real, verifiable sources — just like Jenni.ai and Elicit.
How Grounded Generation Works
- You prompt — Write what you need (essay, research paper, literature review)
- Ntekuma searches — Queries 3 academic databases in parallel for real papers
- Full text retrieved — Fetches abstracts + full-text from open access and your subscribed databases
- Grounded generation — The AI writes your text using actual findings from the papers
- Auto-citation — Inline (Author, Year) citations and a formatted References section
- Verification — Claims are checked against source data for accuracy
Academic Databases
Credential Vault (Access Gated Content)
Connect your publisher API keys and institutional access to retrieve full-text articles from paywalled databases. All credentials are AES-256-GCM encrypted and never stored in plaintext.
Manage your credentials at /api/credentials. Setup instructions are provided for each database.
AI Detection Evasion
All generated text passes through a 15-step humanization pipeline targeting 85+ detection scores on GPTZero, Turnitin, and Originality.ai. The system uses Orwell-style writing rules, perplexity boosting, burstiness variation, and AI-phrase elimination.
Troubleshooting
Preview is blank
Check that the entrypoint file (index.html) exists
Output looks wrong
Narrow your prompt and specify a single target
Build fails
Review the error logs and confirm model access
Build drifts from original
Use explicit 'keep this, change that' edits
Slow generation
Try a faster model for iterations, use Pro models for final
Need More Help?
Contact our support team at support@ntekuma.com or check our community forum for tips and examples.