Orizon QA supports four AI provider options. You configure your provider once in Settings → AI Services, and it applies to all analysis, agent testing, and assistant features.
All API keys are encrypted with AES-256-GCM before being stored. Your keys are never stored in plain text.
Opening AI settings
Go to Settings
Click Settings in the sidebar navigation.
Select AI Services
Choose the AI Services tab to see all provider configuration options.
Save your configuration
Enter your credentials for the provider you want to use and click Save.
Providers
Anthropic
OrizonLLM / LiteLLM
Ollama
LM Studio
Connect directly to Anthropic’s Claude API.To configure:
- Go to console.anthropic.com and create an API key.
- In Settings → AI Services, paste your key into the Anthropic API Key field.
- Click Save.
Anthropic is used as the fallback provider when OrizonLLM/LiteLLM is not configured. OrizonLLM is a multi-provider gateway built on LiteLLM. It routes requests to Anthropic, OpenAI, Azure, Google, and other providers through a single unified API.To configure:
- In Settings → AI Services, enable OrizonLLM.
- Enter your endpoint URL (e.g.,
https://your-litellm-instance.com).
- Enter your API key for the gateway.
- Set your default model using the
provider/model format.
Model format examples:anthropic/claude-3-5-haiku-20241022
openai/gpt-4o
azure/gpt-4o-deployment
OrizonLLM is the recommended option if you want to switch between providers without re-entering credentials, or if you need to route traffic through your own infrastructure.
Run models locally or on a self-hosted server using Ollama.To configure:
- Install and start Ollama on your machine or server.
- In Settings → AI Services, enter your Ollama URL (default:
http://localhost:11434).
- Click Save.
Ollama supports any model available through its registry, including llama3, mistral, and others you have pulled locally.Ollama does not require an API key when running locally. If your Ollama instance requires authentication, enter the key in the Ollama API Key field.
Run models locally using LM Studio.To configure:
- Start LM Studio and load a model.
- Enable the local server in LM Studio (default endpoint:
http://localhost:1234).
- In Settings → AI Services, enter your LM Studio URL.
- Click Save.
LM Studio does not require an API key. Leave the API key field empty when using this provider. Key security
When you save an API key, Orizon QA encrypts it using AES-256-GCM before writing it to the database. The encryption key is stored separately from your data. Your plain-text API key is never persisted anywhere.
If you need to update a key, enter the new value and save — the old encrypted value is replaced. If you leave the field empty when saving other settings, the existing key is preserved unchanged. 