AI Models & Selection
CodeEx AI supports 13+ specialized models from multiple providers. Learn how to choose the right model for your tasks.
Auto Mode (Recommended)
Let CodeEx AI automatically select the best model for each query
When set to "Auto", CodeEx AI analyzes your query and routes it to the most appropriate model:
🧮 Math Problems
Routes to math-specialized models
💻 Code Questions
Uses coding-optimized models
🖼️ Image Tasks
Selects multimodal models
💬 General Chat
Picks conversational models
Model Categories
General Purpose
Versatile models for everyday tasks and conversations
Llama 3.1 8B Instant
Groq • Fast and efficient for general tasks
Llama 3.2 1B Instruct (Free)Free
HuggingFace • Compact and efficient - completely free
Llama 3.1 70B Instruct (Free)Free
HuggingFace • Large and powerful - completely free
Qwen 2.5 7B Instruct (Free)Free
HuggingFace • Strong multilingual capabilities - completely free
Coding Specialists
Models optimized for programming and software development
DeepSeek V3.2
HuggingFace • Advanced coding and reasoning model
Conversational
Models fine-tuned for natural dialogue and chat
RNJ-1 Instruct
HuggingFace • Efficient conversational AI model
Gemini 2.5 Flash Lite
Google • Lightweight conversational model
Multimodal (Vision)
Models that can understand and process images
Gemini 2.5 Flash
Google • Latest multimodal model with vision capabilities
Gemini Flash Latest
Google • Latest version with improved capabilities
Free Models
Completely free models with no usage limits
OpenRouter Auto (Free)Free
OpenRouter • Automatically selects best free model
How to Select a Model
Step-by-step guide to choosing and switching models
Desktop
- 1.Click the Settings icon (⚙️) in the header
- 2.Find the "AI Model" dropdown
- 3.Select "Auto" or choose a specific model
- 4.Settings are saved automatically
Mobile
- 1.Tap the model button in the chat interface
- 2.A bottom sheet opens with all models
- 3.Tap to select your preferred model
- 4.The sheet closes and model is applied
💡 Model Selection Tips
Choose the right model for better results
For Coding Tasks
Use DeepSeek V3.2 for complex programming questions, debugging, and architecture advice.
For Math Problems
Auto mode automatically routes to math-capable models like Qwen 2.5 or Llama models.
For Image Analysis
Use Gemini models for image understanding, visual Q&A, and multimodal tasks.
For General Chat
Auto mode or conversational models like RNJ-1 work best for everyday conversations.
For Free Usage
Use HuggingFace free models or OpenRouter Auto for unlimited usage without costs.
For Speed
Groq models (Llama 3.1 8B Instant) provide the fastest response times.
Model Providers
Groq
Ultra-fast inference with streaming support
HuggingFace
Open-source models with free options
Advanced multimodal capabilities
OpenRouter
Aggregated access to multiple models