Skip to main content

Model Configuration

PostQode supports multiple AI providers and models, giving you flexibility to choose the best option for your needs. You can configure models for both Chat and Agent modes.

Accessing Model Configuration​

  1. Open VS Code with PostQode extension installed
  2. Click on the PostQode icon in the sidebar to open the PostQode panel
  3. Click on the Settings icon (gear icon) in the PostQode panel
  4. Navigate to the Model Configuration section

Operating Modes​

PostQode offers two distinct modes for different use cases:

Chat Mode​

  • Interactive conversational interface
  • Best for quick questions and discussions
  • Lightweight model usage
  • Real-time responses

Agent Mode​

  • Advanced task execution capabilities
  • Tool usage and file operations
  • Complex workflow automation
  • Extended context handling

Supported API Providers​

PostQode supports a wide range of AI providers:

Premium Providers​

  • PostQode - Integrated provider with optimized models
  • Anthropic - Claude models with advanced reasoning
  • OpenAI - GPT models with various capabilities
  • Google Gemini - Google's multimodal AI models
  • Amazon Bedrock - Enterprise-grade AI models
  • GCP Vertex AI - Google Cloud AI platform

Budget-Friendly Options​

  • OpenRouter - Access to multiple models through one API
  • Groq - High-speed inference
  • DeepSeek - Cost-effective models
  • Together - Open-source model hosting
  • Fireworks AI - Fast inference platform

Local and Self-Hosted​

  • Ollama - Run models locally
  • LM Studio - Local model management
  • LiteLLM - Universal LLM proxy

OpenRouter Configuration​

OpenRouter provides access to multiple AI models through a single API, making it an excellent choice for cost-conscious users and those wanting model variety.

Setting Up OpenRouter​

  1. Get API Key: Visit OpenRouter to obtain your API key
  2. Configure in PostQode:
    • Select "OpenRouter" from the API Provider dropdown
    • Enter your OpenRouter API Key
    • Choose from available models

OpenRouter Benefits​

  • Cost Effective: Access to free and low-cost models
  • Model Variety: Choose from dozens of different models
  • No Vendor Lock-in: Switch between models easily
  • Transparent Pricing: Clear cost per token pricing

Free Models Available​

OpenRouter offers several free models that are perfect for getting started:

  • Free Tier Models: Various open-source models with no cost
  • Rate Limited: Some models have usage limits but are free
  • Community Models: Models supported by the community

Anthropic Configuration​

API Key Setup​

  1. Obtain your Anthropic API key from the Anthropic Console
  2. Enter the key in the "Anthropic API Key" field
  3. The key is stored locally and only used for API requests

Model Features​

Note: The following features are examples from Anthropic Claude models. Available features vary significantly between different models and providers.

Context Window Options​

  • Standard: 1,000,000 tokens for most models
  • 200K Context: Switch to 200K context window model for faster processing

Advanced Features​

  • Extended Thinking: Enable for complex reasoning tasks
  • Image Support: Process and analyze images
  • Browser Use: Web interaction capabilities
  • Prompt Caching: Reduce costs with cached prompts

Pricing Structure​

Note: The pricing shown below is an example from Anthropic Claude models. Pricing varies significantly between providers and models. Always check your provider's current pricing before use.

Input Pricing (per million tokens)​

  • ≤ 20,000 tokens: $3.00/million tokens
  • > 20,000 tokens: $6.00/million tokens

Cache Pricing​

  • Cache Writes: $3.75/million tokens
  • Cache Reads: $0.30/million tokens

Output Pricing (based on input tokens)​

  • ≤ 20,000 tokens: $15.00/million tokens
  • > 20,000 tokens: $22.50/million tokens

Custom Configuration Options​

Base URL Override​

  • Use custom base URL: Enable for self-hosted or proxy endpoints
  • Useful for enterprise deployments or custom routing

Model Selection​

  • Choose from available models in the dropdown
  • Models are filtered based on selected provider
  • Real-time availability and pricing information

API Key Management​

  • Store keys securely in PostQode settings
  • Keys are encrypted and stored locally
  • Never share API keys in code or documentation
  • Rotate keys regularly for security
Is this documentation incorrect or incomplete? Report an issue on GitHub