AI Providers
Supported Providers
Section titled “Supported Providers”Cortiq is designed to work with the main provider families used across the product, including OpenAI ChatGPT, Google Gemini, Anthropic Claude, and xAI Grok.
The exact integration route exposed to the user depends on the provider and the current application surface. In the current desktop settings screen, the main distinction presented to the user is:
- direct API integration
- local tool or CLI integration where supported
Who This Is For
Section titled “Who This Is For”This page is for users who want to:
- choose the right provider for their workflow
- understand the difference between local-tool and API-based operation
- decide whether traceability, convenience, or fallback behavior matters most
- avoid choosing a provider setup that does not match their operating style
Where To Find This In The App
Section titled “Where To Find This In The App”Provider setup lives mainly in:
Settings->AI Providers
That screen is where you configure provider credentials, choose API or CLI/browser-style integration where supported, and set feature-level provider overrides.
Per-session provider choice then happens when you build or edit a session in:
Library->Sessions
How The Settings Screen Labels This
Section titled “How The Settings Screen Labels This”In the current customer UI, provider integration choices are surfaced as API and CLI where a provider supports both.
For customers, the important practical distinction is simple:
APImeans Cortiq is calling the provider directly through an API key and structured adapter.CLImeans Cortiq is using a local-tool route on the same machine instead of a direct API call.
Two Integration Modes
Section titled “Two Integration Modes”Local Tool Or CLI Route
Section titled “Local Tool Or CLI Route”The local-tool route uses a provider-compatible tool path on the same machine instead of direct API billing.
This mode is useful when you want:
- a provider workflow that does not depend on the same direct API setup as the API route
- a local-machine integration path where the provider tool handles the interaction
- flexibility when your preferred provider workflow is better suited to a local tool than to a straight API call
API Mode
Section titled “API Mode”API mode uses direct provider APIs through Cortiq adapters.
This mode is useful when you want:
- Structured request and response handling
- Persisted conversation history
- Token usage tracking
- Fallback provider behavior if the primary provider fails
- Cleaner automation for serious operational usage
How Cortiq Chooses The Route
Section titled “How Cortiq Chooses The Route”Each session stores both an AI provider and an integration type. Cortiq then routes the session to the matching local-tool or API pipeline automatically.
That means one session can run with one provider and mode, while another session uses a different provider and mode entirely.
Feature-Level Provider Overrides
Section titled “Feature-Level Provider Overrides”The Settings -> AI Providers screen also includes Feature AI Providers.
That area lets you assign a specific provider and integration type to a particular Cortiq feature instead of inheriting the session default.
This is useful when:
- one provider is better for your main trading loop but another is better for a supporting workflow
- you want to keep the session default simple but tune one advanced feature separately
- you need to isolate a weaker provider from the rest of the product surface
Security And Credentials
Section titled “Security And Credentials”Cortiq stores supported provider API keys in encrypted local storage on Windows.
For local-tool routes, the main setup requirement is the local executable path and related configuration rather than an API key alone.
Practical Selection Guide
Section titled “Practical Selection Guide”| If You Need | Prefer |
|---|---|
| A local-machine route where direct API setup is not your first choice | Local tool or CLI route |
| Better traceability and cleaner operational control | API mode |
| Provider fallback during runtime | API mode |
| Feature-level provider overrides | Settings -> AI Providers |
Feature Implications
Section titled “Feature Implications”Your provider choice affects more than response text. It also affects:
- Session traceability
- Conversation persistence
- Operational cost structure
- Failure handling options
- Setup effort for first use
Best Use Cases
Section titled “Best Use Cases”Choose this part of the platform carefully when you want:
- lower-friction experimentation with provider web apps
- a cleaner API-driven operating model for serious runtime usage
- different providers for different session types
- resilience through fallback-provider design