Get Your Development Environment Ready
Follow these steps to set up your coding environment and claim your AI agent
🎥 Watch the Setup Demo Video
2 Minute Setup Video (Please use it only as a setup reference, not for ideation. What you build is completely up to your creativity!)
⚡ Quick update gooo
Watch Setup Demo on YouTube💡 Beginner-Friendly Recommendations
•Easiest Option: Choose Web SDK - runs in browser, no mobile setup needed!
•For Windows Users New to Mobile: Choose Kotlin - just hit Run in Android Studio
•For Coding Agent: Use OpenCode Desktop for a visual interface (download after CLI setup)
1. Choose Your Tech Stack
Select the technology you want to use for your project. This will customize the setup instructions for you.
2. Select Your Development Platform
Choose your development machine (Windows/Mac/Linux). Installation instructions will adapt to your selection.
Please select a tech stack first to see available platforms.
3. Select Your Target Device
Choose which device you'll be running your app on (Android or iOS).
Please select a tech stack first to see available target devices.
4. Follow the Setup Steps
Follow these steps to set up your environment. Checkboxes are optional - just for tracking your progress.
Step 1: Get Your API Key
To get started, you need an API key to access the AI coding agent.
**Option A: API Key Emailed to You** 📧
**Check your email first!** Your API key should have been emailed to you. If you received it, save it and skip to Step 3.
Option B: If Not Emailed, Register via Website
If you didn't receive an email, follow these steps:
1. Go to the registration page: https://claim-coding-agent.vercel.app/register
2. Sign in with your GitHub account
3. Complete the registration form
4. Wait maximum 24 hours before an admin assigns you your secret key
5. Find your key at: https://claim-coding-agent.vercel.app/mykeys
⚠️ Important: Use the same email address as registered in the hackathon!
Step 2: Save Your API Key
Now that you have your API key (either from email or the website), make sure to save it safely.
If you got your key via email:
📧 Copy your RUNANYWHEREAI_KEY from the email (looks like: sk-xxxxxxxxxxxxx)
If you registered via website:
1. Go to: https://claim-coding-agent.vercel.app/mykeys
2. Sign in with the same GitHub account
3. Copy your RUNANYWHEREAI_KEY (looks like: sk-xxxxxxxxxxxxx)
Save this key - you'll need it in the next step!
💡 Tip: You have limited prompts, so use them wisely! Choose the right model for your task:
• Claude Sonnet 4.5: ~50-80 prompts (use for complex tasks, reasoning, code generation)
• Claude Haiku 4: ~200-300 prompts (use for simple tasks, quick questions, testing)
Need more credits? Submit your mid-week checkpoint to request additional budget!
Step 3: Install OpenCode CLI
👆 Select your platform above to see installation instructions.
Select your platform to see installation instructionsStep 4: Install OpenCode Desktop (Recommended)
💡 Much Easier Than Terminal!
OpenCode Desktop provides a user-friendly GUI interface. Since you ran the CLI install in Step 3, your API key is already configured!
1. Download: https://opencode.ai/download
2. Install and launch the OpenCode Desktop app
3. **No API key prompt** - it automatically uses your existing key
If you don't see RunAnywhere models:
1. Click the **Models button** (below the chatbox)
2. Click **Manage Models**
3. Add RunAnywhere models under the **RunAnywhereAI** section
If models still don't appear:
Re-run the installation commands from Step 3, then restart OpenCode Desktop.
To change your API key:
Run the installation commands from Step 3 again with your new key.
Step 5: Fork & Clone Starter Repository
Select your tech stack above to see specific instructions for forking and cloning the starter repository.
Step 6: Setup Your IDE
Select your tech stack above to see specific IDE setup instructions.
Step 7: Setup Your Test Device
Select your tech stack and target device above to see device setup instructions.
Step 8: Run Your First Build
Select your tech stack above to see specific run instructions.
Select your tech stack to see run instructionsStep 9: Give Documentation to Your Coding Agent
Start your coding agent and give it the RunAnywhere documentation:
1. Open your terminal in the project folder
2. Run: opencode
3. When the agent starts, paste this message:
"I'm building a mobile app using RunAnywhere SDK. Please refer to https://docs.runanywhere.ai/ for all RunAnywhere documentation. Tell me your tech stack. The starter app already has RunAnywhere SDK integrated. Help me implement features using the on-device LLM, STT, TTS, and VAD capabilities."
4. Now ask the agent to help you implement your idea!
Available RunAnywhere SDK features:
• LLM: On-device AI text generation
• STT: Speech-to-Text conversion
• TTS: Text-to-Speech synthesis
• VAD: Voice Activity Detection
✅ Your agent now knows about RunAnywhere SDK and can help you implement features using the documentation.
🆘 Need Help?
• Join Discord office hours: https://discord.gg/N359FBbDVd
• RunAnywhere SDK: https://github.com/RunanywhereAI/runanywhere-sdks
• Full documentation: https://docs.runanywhere.ai/