Custom AI Integration
Configure your own AI provider for better control, cost optimization, and compliance with company policies.
Why Use Custom AI Integration?
By default, Reqase Lite uses a built-in AI service. However, you can configure your own AI providers for:
Better Control
Choose specific AI models and configure parameters to match your testing requirements
Cost Optimization
Use your existing AI subscriptions and manage costs according to your budget
Compliance
Meet company policies by using approved AI providers and keeping data within your infrastructure
Specific Capabilities
Access specialized AI models optimized for your domain or testing scenarios
Supported AI Providers
1. OpenAI
Models: GPT-4, GPT-4 Turbo, GPT-3.5 Turbo
API Key: Get from OpenAI Platform
Best for: General-purpose test generation with high-quality outputs
2. Google Gemini
Models: Gemini 1.5 Pro, Gemini 1.5 Flash
API Key: Get from Google AI Studio
Best for: Fast generation with good context understanding
3. Anthropic Claude
Models: Claude 3.5 Sonnet, Claude 3 Opus, Claude 3 Haiku
API Key: Get from Anthropic Console
Best for: Detailed test scenarios with strong reasoning capabilities
4. DeepSeek
Models: DeepSeek Chat, DeepSeek Coder
API Key: Get from DeepSeek Platform
Best for: Technical test cases and API testing scenarios
5. Custom LLM
Requirements: Must support OpenAI API-compatible endpoints
Use Case: Self-hosted models or other compatible providers
Compatible with: LiteLLM, LocalAI, FastChat, vLLM, and similar frameworks
Adding a Custom AI Provider
Go to Settings → Custom AI Integration
Click Add New API button
Fill in the configuration:
- AI Provider: Select from dropdown
- Model: Choose the specific model
- API Key: Enter your API key
- For Custom LLM:
- API URL: Enter the endpoint URL
- Model Name: Enter the model identifier
Click Test Connection to verify
Click Save once the test succeeds
Toggle the switch to enable the provider
Managing Multiple Providers
Add Multiple
You can add multiple AI providers and switch between them as needed
One Active
Only one provider can be enabled at a time for test generation
Edit & Delete
Click the edit icon (✏️) to modify or delete icon (🗑️) to remove providers
Fallback
If all custom providers are disabled, the system uses the built-in AI
Next Steps
With your AI provider configured, you can learn more about custom LLM requirements or continue with project settings.