Overview

WatchThatSite lets you connect your existing AI providers like OpenAI, Anthropic, Google, and more. You can add your existing api keys, configure allowed models and you're good to go. If you add more than one key for the same model, we will distribute traffic among your keys for that model.

All major providers are supported. If you're favourite provider isn't, let us know!

How it works

1. You pick a provider and add your credentials (all your credentials are stored securely).
2. You configure your presets with model and prompts.
3. Your monitors will use your presets.

You are not locked in to any provider or model. You can switch providers or models at any time and experiment what works best for you. You can even use self-hosted inference engines like Ollama, VLLM, etc.

Security

Your credentials are encrypted using industry-standard AES-256 using Argon2id. Your credentials are never logged or cached unencrypted. We never access your provider account except when you explicitly request an operation (e.g., listing models, making inference calls).

Available providers

Below is a list of providers we support.

• Anthropic
• Azure
• Bedrock
• Cerebras
• Cohere
• Fireworks
• Gemini
• Groq
• Index
• Meta
• Mistral
• Ollama
• OpenAI
• Openrouter
• Vertex
• VLLM
• xAI

If there's a provider you are using and is not supported yet, please let us know 🙇‍♂️

Available models

Almost all models from each provider are supported. Only criteria is they need to support tool calling or function calling.

Managing provider keys

Add a provider key

1. Go to Settings > AI Providers
2. Click Add Provider Key
3. Select your provider from the dropdown
4. Enter required credentials (API key, endpoint, region, etc.)
5. Optionally select specific models to restrict access (leave empty to allow all models)
6. Click Save

Edit a provider key

1. Go to Settings > AI Providers
2. Click Edit next to the provider you want to modify
3. Update credentials (blank fields keep current values on edit)
4. Click Save

Enable or disable a provider key

Toggle the switch next to any provider key to enable or disable it without removing your encrypted credentials.

Delete a provider key

Click Delete next to a provider key to permanently remove it and its stored credentials. This action cannot be undone.

Supported providers

Cloud providers (API Key)

Simple setup with just an API key:

• OpenAI
• Anthropic
• Google Gemini
• Groq
• Mistral
• Cohere
• xAI
• OpenRouter
• Fireworks
• Cerebras

Enterprise providers (multi-credential)

Complex authentication with endpoints, regions, or service accounts:

• Azure OpenAI - API key + endpoint + API version
• AWS Bedrock - Access key + secret + region
• Google Vertex AI - Project ID + service account JSON + region

Self-Hosted

Run models on your own infrastructure:

• Ollama - Simple local LLM server
• vLLM - High-throughput inference engine

Security best practices

• Rotate regularly: Rotate your API keys regularly
• Monitor usage: Check your provider's usage dashboard for unusual activity
• Use dedicated keys: Create API keys specifically for WatchThatSite (don't reuse main account keys)
• Endpoint security: For self-hosted models (Ollama, vLLM), use network isolation or VPN