Sign in
Enter your credentials to access EDYA
Untitled chat
Admin mode
⚠ Pending from Developer
Prompt
EDYA
Welcome. Describe the app you want to build and I'll generate it live in the preview.
Todo app BMI calc Pomodoro Palette gen
Ready
⚡
Preview appears here
Type a prompt and hit send

    

    

     No app generated yet
     Web view
    

   

  

 


 
 

  

   

    
All chats
Your conversation history with EDYA

    

     
     
    

   

   

    
Total chats
0
conversations saved

    
This week
0
new chats

    
Apps generated
0
across all chats

    
Active users
1
chatting this session

   

   

    

     Conversations
     

      
     

    

    

   

  

 


 
 

  

   

    
Users & Access
Manage team members, roles and permissions

    
   

   

    
Total users
—
in this workspace

    
Online
—
active recently

    
Admins
—
can commit to GitHub

    
Revoked
—
access removed

   

   

    
Team members

    
UserRolePortal Dev2FAActions

   

  

 


 
 

  

   
Activity Log
Full audit trail of every action in your workspace

   

    

     All activity
     

      
      
      
     

    

    
TimeUserActionDetail

   

  

 


 
 

  

   

    

     
Server Management

     
Configure GPU inference providers and routing priorities — admin only

    

    

     
     
    

   

   

    
Total servers
0
configured providers

    
Active
0
serving requests

    
Healthy
0
passing health check

    
Offline
0
need attention

   

   

    

     Inference Providers
     

      
      
     

    

    

     

      

       #
       NameProviderPriority
       StatusHealthModels
       Last checkedActions
      

      
     

    

   

   

    Routing logic: Requests route to the lowest-priority active+healthy server supporting the requested model. Equal-priority servers are load-balanced round-robin. Offline or inactive servers are skipped automatically.
   

  

 








 

  

   
Add server
Configure a GPU inference provider endpoint

   
  

  

   

   

    

    

     
    

   

   

   

    

     
     
    

   

   

    

    

   

   

    
Active
Inactive servers are skipped by the router

    
   

  

  

   
   
   

   
   
  

 







 

  
Delete server
This cannot be undone

  
You are about to permanently delete this server. All configuration will be lost.

  

   
   
  

 





 

  

   

    

     
API Connectors

     
Connect AI services, open-source models and external APIs — admin only

    

    
   

   


    
API Connectors
Connect AI services, open-source models and external APIs — admin only

    

    
    
Anthropic

    

     

      
      

       
Anthropic Claude Latest

       
Opus 4 · Sonnet 4.6 · Haiku 4.5 — EDYA's native AI engine

      

      Not connected
      
     

     

      
claude-opus-4-6claude-sonnet-4-6claude-haiku-4-5

      

       

       

        
        
       

      

      

     

    


    
    
OpenAI

    

     

      
      

       
OpenAI Popular

       
GPT-4o · GPT-4o mini · o1 · o3 — OpenAI API

      

      Not connected
      
     

     

      
gpt-4ogpt-4o-minio1o3gpt-3.5-turbo

      

       

       

        
        
       

      

      
Base URL (optional override)

      

     

    


    
    
Google

    

     

      
      

       
Google Gemini 2.0

       
Gemini 2.0 Pro · Flash · Ultra — Google AI Studio

      

      Not connected
      
     

     

      
gemini-2.0-progemini-2.0-flashgemini-1.5-progemini-1.5-flash

      

       

       

        
        
       

      

      

     

    


    
    
xAI

    

     

      
      

       
xAI Grok

       
Grok 3 · Grok 3 Mini — xAI API

      

      Not connected
      
     

     

      
grok-3grok-3-minigrok-2

      

       

       

        
        
       

      

      

     

    


    
    
Mistral AI

    

     

      
      

       
Mistral AI Open

       
Mistral Large · Small · Codestral — La Plateforme API

      

      Not connected
      
     

     

      
mistral-large-latestmistral-small-latestcodestral-latestopen-mixtral-8x22b

      

       

       

        
        
       

      

      

     

    


    
    
Cohere

    

     

      
      

       
Cohere

       
Command R+ · Command R — Enterprise RAG-optimised

      

      Not connected
      
     

     

      
command-r-pluscommand-rcommand

      

       

       

        
        
       

      

      

     

    


    
    
Open Source · Self-hosted


    

     

      
      

       
Groq Fast

       
LPU inference — Llama 3 · Mixtral · Gemma at 300+ tok/s

      

      Not connected
      
     

     

      
llama-3.3-70bllama-3.1-8bmixtral-8x7bgemma2-9b

      

       

       

        
        
       

      

      

     

    


    

     

      
      

       
Together AI OSS

       
50+ open-source models — DeepSeek · Llama · Qwen

      

      Not connected
      
     

     

      
deepseek-r1Llama-3.3-70BQwen2.5-72BWizardLM-2-8x22B

      

       

       

        
        
       

      

      

     

    


    

     

      
      

       
Hugging Face Hub

       
Inference API · Serverless · 500k+ models

      

      Not connected
      
     

     

      
meta-llama/Llama-3.3-70Bmistralai/Mistral-7Bgoogle/gemma-2-9bdeepseek-ai/DeepSeek-R1

      

       

       

        
        
       

      

      

       
Pull model directly from HuggingFace Hub

       

        
        
       

       

      

     

    


    

     

      
      

       
Replicate

       
Run open-source models in the cloud — pay per second

      

      Not connected
      
     

     

      
meta/llama-3.3-70bmistralai/mixtral-8x7bstability-ai/sdxl

      

       

       

        
        
       

      

      

     

    


    

     

      
      

       
Ollama Local

       
Self-hosted local inference — run any GGUF model on your machine

      

      Not connected
      
     

     

      
llama3.2mistraldeepseek-r1qwen2.5-coderphi4

      
Ollama server URL

      

       

        
        
       

      

      

     

    


    
    
GitHub

    

     

      
      

       
GitHub Models

       
Access frontier models via GitHub token — free tier available

      

      Not connected
      
     

     

      
gpt-4oMeta-Llama-3.1-70BPhi-3.5-miniMistral-large

      

       

       

        
        
       

      

      

       
Pull model weights from a GitHub repo

       

        
        
       

       

      

      

     

    


    
    
Other & Custom


    

     

      
      

       
DeepSeek R1

       
DeepSeek-R1 · V3 — reasoning + coding specialist

      

      Not connected
      
     

     

      
deepseek-r1deepseek-v3deepseek-coder-v2

      

       

       

        
        
       

      

      

     

    


    

     

      
      

       
Perplexity AI

       
Sonar models with live web search built-in

      

      Not connected
      
     

     

      
sonar-prosonarsonar-reasoning

      

       

       

        
        
       

      

      

     

    


    

     

      
      

       
Custom OpenAI-compatible endpoint

       
Any vLLM, LiteLLM, LocalAI, or OpenAI-compatible server

      

      Not connected
      
     

     

      
Base URL *

      

       

       

        
        
       

      

      
Model name

      

     

    


    
    
Open Source LLM Import

    

      Pull or upload any open-source LLM. Imported models appear automatically in the model selector below the prompt box. Supports GitHub repos, Hugging Face model IDs, and direct file uploads (.gguf, .bin, .onnx, .pt).
    


    
    

     

      
      

       
Import from GitHub Repo

       
Pull any open-source LLM codebase or weights from a public GitHub repo

      

      
     

     

      

       
       

        
        
       

      

      

       
       
      

      

     

    


    
    

     

      
      

       
Upload model file Local

       
Upload a .gguf, .bin, .onnx or .pt model file from your machine

      

      
     

     

      

       
       
🧠

       
Drop model files here or click to browse

       
Supports .gguf · .bin · .onnx · .pt · .safetensors · .ckpt

      

      

     

    


    
    

   

  

 





 

  
Settings
Profile, workspace, security and integrations

  

   

    
    
    
    
    
    
    
    
    
    
    
    

    

   
   

   

   
   

   
   

   
   


   
   

    
Your profile
How you appear to your team

    

     
AK

     

      
      
     

    

    

     

     

    

    

    

    

   


   
   

   
   

   
   

   
   

   
   

   
   

  

 







 
Invite developer
They'll receive an email with 2FA setup instructions

 
2FA is mandatory. Developers must complete authenticator setup before accessing EDYA.

 







 
Submit changes
Admin will review before changes go live

 
Not live until an admin approves and commits to GitHub.

 







 
Commit to GitHub
Only changed sections will be pushed — not the full file

 

  
✓ Analysing changes…

  

   

  

  

  

 

 







 
Log out
You'll need to sign in again

 
Any unsaved work will be lost.