AI Provider Configuration

Configure OpenAI, Azure AI, or OpenRouter for VISTA’s AI capabilities.

Table of contents
  1. Supported Providers
  2. OpenAI Configuration
    1. Getting an API Key
    2. Settings
    3. Recommended Models
  3. Azure AI Configuration
    1. Endpoint Types
      1. Azure AI Foundry (New — Recommended)
      2. Classic Azure OpenAI
    2. Auth Modes
      1. API Key (Default)
      2. Azure Identity (Keyless / Entra ID)
    3. API Version
    4. Settings Summary
  4. OpenRouter Configuration
    1. Getting an API Key
    2. Settings
    3. Recommended Free Models
  5. Advanced Settings
    1. Temperature
    2. Max Tokens
    3. Connection Testing
  6. Configuration Storage
  7. Switching Providers

Supported Providers

Provider Models Cost Best For
OpenAI GPT-4o, GPT-4o-mini, GPT-5, o1/o3/o4 Pay-per-use Direct API, latest models
Azure AI Any deployed model incl. GPT-5, o-series Pay-per-use Enterprise, compliance, Azure Identity
OpenRouter 500+ models, free tier available Free + paid Everyone — fastest way to start

OpenAI Configuration

Getting an API Key

  1. Go to platform.openai.com
  2. Sign up or log in
  3. Navigate to API Keys
  4. Click “Create new secret key”
  5. Copy the key (starts with sk-...)

Settings

Setting Value
Provider OpenAI
API Key sk-...
Model gpt-4o-mini (recommended for cost)
Model Context Cost Notes
gpt-4o-mini 128K ~$0.001/req Cost-effective, fast — good default
gpt-4o 128K ~$0.01/req Higher quality analysis
gpt-5.2 128K varies Latest GPT-5 family
o4-mini 128K varies Reasoning model — temperature omitted automatically
o3 128K varies Reasoning model — temperature omitted automatically

VISTA automatically detects o1, o3, and o4-series models and omits the temperature parameter, which these models reject. No manual configuration needed.

Azure AI Configuration

VISTA supports two Azure endpoint flavours and two auth modes, auto-detected from your endpoint URL.

Endpoint Types

Azure AI Foundry (New — Recommended)

Endpoints on *.services.ai.azure.com:

Endpoint:   https://your-resource.services.ai.azure.com/openai/v1
Deployment: gpt-5.2   (or any deployed model name)

VISTA sends the model name in the request body, matching the Azure AI Foundry OpenAI-compatible API format.

Classic Azure OpenAI

Endpoints on *.openai.azure.com:

Endpoint:   https://your-resource.openai.azure.com
Deployment: your-deployment-name
API Version: 2025-01-01-preview  (configurable)

VISTA puts the deployment name in the URL path, matching the classic Azure OpenAI format.

VISTA auto-detects the endpoint type from your URL — no manual selection needed. Just paste your endpoint and it routes correctly.

Auth Modes

API Key (Default)

Standard key-based authentication:

Setting Value
Provider Azure AI
Auth Mode API Key
API Key Your Azure key
Endpoint Your endpoint URL
Deployment Your deployment / model name

Finding your credentials:

  • Azure Portal → Your AI resource → Keys and Endpoint
  • Copy Key 1 or Key 2 and the Endpoint URL
  • Deployment name: Azure AI Foundry Portal → Deployments tab

Azure Identity (Keyless / Entra ID)

Authenticate without an API key using Microsoft Entra ID. VISTA implements the DefaultAzureCredential chain in pure Java:

  1. Environment variables — AZURE_TENANT_ID + AZURE_CLIENT_ID + AZURE_CLIENT_SECRET
  2. Azure CLI — az login (fallback; runs az account get-access-token)
Setting Value
Provider Azure AI
Auth Mode Azure Identity
Endpoint Your endpoint URL
Deployment Your deployment / model name
API Key (leave blank — not needed)

Azure Identity setup: Set AZURE_TENANT_ID, AZURE_CLIENT_ID, and AZURE_CLIENT_SECRET in your environment before launching Burp Suite, or run az login in a terminal. Tokens are cached and refreshed automatically.

API Version

The API version is configurable in Settings (default: 2025-01-01-preview). Only applies to Classic Azure OpenAI endpoints — Foundry endpoints don’t need it.

Settings Summary

Setting Classic Azure OpenAI Azure AI Foundry
Endpoint https://xxx.openai.azure.com https://xxx.services.ai.azure.com/openai/v1
Deployment Deployment name from Studio Model name (e.g. gpt-5.2)
API Version 2025-01-01-preview N/A (auto)
Auth Mode API Key or Azure Identity API Key or Azure Identity

OpenRouter Configuration

Getting an API Key

  1. Go to openrouter.ai
  2. Sign up — no credit card required for free models
  3. Visit openrouter.ai/keys
  4. Click “Create Key”
  5. Copy the key (starts with sk-or-v1-...)

Settings

Setting Value
Provider OpenRouter
API Key sk-or-v1-...
Model deepseek/deepseek-r1-0528:free
Model ID Quality Speed
DeepSeek R1 0528 ⭐ deepseek/deepseek-r1-0528:free ⭐⭐⭐⭐⭐ Medium (reasoning)
Llama 3.3 70B meta-llama/llama-3.3-70b-instruct:free ⭐⭐⭐⭐⭐ Fast
DeepSeek R1T2 Chimera tngtech/deepseek-r1t2-chimera:free ⭐⭐⭐⭐⭐ Medium

Free models on OpenRouter rotate periodically. If a model returns a 404, switch to another from the list above.

Advanced Settings

Temperature

Controls response randomness:

Value Behavior Use Case
0.0 Deterministic, focused Consistent analysis
0.3 Slightly creative (default) Balanced testing guidance
0.7 Creative Exploring unusual bypass vectors
1.0 Maximum creativity Brainstorming only

The default temperature of 0.3 is optimal for security testing. For reasoning models (o1, o3, o4-series), temperature is omitted automatically.

Max Tokens

Controls maximum response length (uses max_completion_tokens — the current API standard):

Value Approximate Length
1000 Short, focused response
2000 Standard analysis (default)
4000 Detailed methodology
8000 Comprehensive deep dive

Connection Testing

After configuring any provider:

  1. Click “Test Connection” in Settings
  2. VISTA sends a test request to verify credentials
  3. Success: Status bar shows 🟢 with (Provider / model-name)
  4. Failure: Detailed error message with troubleshooting hints

The status bar model name shows the correct model for each provider — deployment name for Azure AI, OpenRouter model for OpenRouter, and model name for OpenAI.

Configuration Storage

AI configuration is stored at ~/.vista-ai-config.json:

{
  "provider": "Azure AI",
  "openaiApiKey": "",
  "azureApiKey": "your-azure-key",
  "model": "gpt-4o-mini",
  "endpoint": "https://your-resource.services.ai.azure.com/openai/v1",
  "deployment": "gpt-5.2",
  "azureApiVersion": "2025-01-01-preview",
  "azureAuthMode": "api-key",
  "openRouterApiKey": "",
  "openRouterModel": "deepseek/deepseek-r1-0528:free",
  "temperature": 0.3,
  "maxTokens": 2000
}

The API key is stored in plaintext in this file. Ensure appropriate file permissions on shared systems. Consider using Azure Identity auth mode to avoid storing keys on disk.

Switching Providers

Switching between providers is instant:

  1. Go to Settings tab
  2. Change the Provider dropdown
  3. Enter the new provider’s credentials
  4. Click “Test Connection”
  5. All VISTA features immediately use the new provider

No restart required. Active conversations continue with the new provider.

Back to top

VISTA — Vulnerability Insight & Strategic Test Assistant. Made with ❤️ for the Security Community.

This site uses Just the Docs, a documentation theme for Jekyll.