AI Providers

SmartCursorX supports multiple AI providers for chat and code completion. Providers are configured in Settings → Models.

Supported Providers

OpenAI

• Models: gpt-4o, gpt-4-turbo, gpt-3.5-turbo, and any custom OpenAI model
• API key format: sk-...
• Base URL can be customized for proxy setups

Anthropic

• Models: claude-3-5-sonnet, claude-3-5-haiku, claude-3-opus
• API key format: sk-ant-...

Gemini

• Models: gemini-2.0-flash, gemini-1.5-flash, gemini-1.5-pro
• API key: From Google AI Studio

OpenRouter

• Access to 200+ models through a single API
• API key format: sk-or-v1-...
• Supports dynamic model discovery and context length lookup
• HTTP-Referer and X-Title headers sent for analytics

Ollama

• Fully local models via localhost:11434
• Requires Ollama server running locally
• Model list fetched dynamically from local Ollama instance
• No API key needed

LiteLLM

• OpenAI-compatible proxy endpoint
• Custom base URL configuration
• Useful for self-hosted model gateways

Zen (Built-in)

• Free tier model via opencode.ai/zen/v1
• No configuration required
• Supports effort levels (low/high)

Local (Experimental)

• Runs quantized GGUF models locally via bundled llama-server
• See Local LLMs for setup
• No API key needed

Fine-Tuned

• User-registered LoRA adapters and fine-tuned models
• Stored in SQLite database
• Inferences run through local endpoints

API Key Security

All API keys are encrypted using OS-level credential storage:

• Windows: DPAPI via Electron safeStorage
• macOS: Keychain
• Linux: libsecret

Keys are never stored as plaintext in database tables or logs. The Secure Store architecture uses an AES-256-GCM encrypted JSON file with a random 32-byte key.

Custom Providers

You can add custom OpenAI-compatible providers:

1. Open Settings → Models
2. Click Add Custom Provider
3. Enter a name, base URL, and API key
4. Models populate automatically from the provider's /v1/models endpoint

Custom Models

Register specific models with custom context lengths and thinking support:

• Set context window size per model
• Toggle reasoning/thinking mode for supported models
• Models are stored in the local SQLite database

Cost Estimation

The Cost Estimator Service tracks token usage per model:

• Input and output token counts
• Per-model pricing (configurable)
• Zero-cost for local, Ollama, and Zen providers
• Usage statistics viewable in Settings → Usage & Costs