Setting Up AI Providers

Before you can use AI Enhancement, you need to connect Echo to an AI provider. This is done in the Settings > Intelligence tab. Once a provider is connected, you enable AI Enhancement per-profile in Adaptive Awareness. Echo supports a wide range of cloud providers as well as local models through Ollama.

Choosing a Provider

Echo supports these AI providers for enhancement:

Provider	Free Tier	Notes
Gemini	Yes	Google's models. Free tier available. Default provider.
Groq	Yes	Very fast inference. Free tier available.
Cerebras	Yes	Fast inference. Free tier available.
OpenAI	No	GPT models. Paid API key required.
Anthropic	No	Claude models. Paid API key required.
Mistral	No	Mistral models. Paid API key required.
OpenRouter	Varies	Access to hundreds of models from many providers. Supports vision models.
Ollama	Yes (local)	Run models on your own Mac. Fully offline, no API key needed.
Custom	Varies	Any OpenAI-compatible API endpoint.

If you're just getting started, Gemini, Groq, or Cerebras are good choices because they offer free API keys.

Adding an API Key

Open Echo and go to Settings
Select the Intelligence tab
Under AI Provider Integration, choose your provider from the AI Provider dropdown
Paste your API key into the API Key field
Click Verify and Save

Echo verifies your key against the provider's API before saving it. If verification fails, double-check that your key is correct and that your account is active.

Once connected, you'll see a green Connected badge next to the provider name.

Where to Get API Keys

Each provider has a link directly in Echo. After selecting a provider, click Get API Key to open the provider's key management page:

OpenAI -- platform.openai.com/api-keys
Anthropic -- console.anthropic.com/settings/keys
Gemini -- makersuite.google.com/app/apikey
Groq -- console.groq.com/keys
Cerebras -- cloud.cerebras.ai
Mistral -- console.mistral.ai/api-keys
OpenRouter -- openrouter.ai/keys

Key Security

Your API keys are stored in the app's preferences.

Selecting a Model

After connecting a provider, you can choose which model to use:

In the Intelligence tab under AI Provider Integration, find the Model dropdown below the provider selector
Select the model you want to use

Each provider offers different models with varying speed, quality, and cost. Faster models (like smaller variants) work well for simple cleanup, while more capable models handle complex rewriting better.

For OpenRouter, click the refresh button next to the model dropdown to load the full list of available models. OpenRouter also supports vision-capable models, which can process screen captures as images rather than text for improved context understanding.

Ollama (Local Models)

Ollama lets you run AI models directly on your Mac with no internet connection and no API key. Everything stays on your device.

Prerequisites

Install Ollama from ollama.ai/download
Open Terminal and pull a model: ollama pull mistral (or any model you prefer)
Keep Ollama running in the background

Connecting Echo to Ollama

Open Echo and go to Settings > Intelligence
Select Ollama from the AI Provider dropdown
The Server URL defaults to http://localhost:11434 -- change it only if you're running Ollama on a different address
Echo will attempt to connect automatically. You should see a green Connected status.
Select your preferred model from the Model dropdown

If the status shows Disconnected, click Refresh to retry the connection. Make sure Ollama is running and that you have at least one model pulled.

Custom Provider

If you use an OpenAI-compatible API that isn't listed (such as a self-hosted model or a third-party service), you can set it up as a custom provider:

Select Custom from the AI Provider dropdown
Enter the Endpoint URL (the full chat completions URL, e.g., https://api.example.com/v1/chat/completions)
Enter the Model name
Enter your API Key
Click Verify and Save

The custom provider uses the OpenAI-compatible API format, so any service that follows that standard should work.

Removing an API Key

To disconnect a provider:

Go to Settings > Intelligence
Select the provider you want to disconnect
Click Remove next to the masked API key

This deletes the key from your Keychain. You can always add a new key later.

Tips

You can add keys for multiple providers and switch between them at any time. Echo remembers your key and model selection for each provider.
Free-tier providers (Gemini, Groq, Cerebras) have rate limits. If you see rate limit errors, wait a moment before trying again or switch to a paid provider.
For the fastest response times, try Groq or Cerebras. For the highest quality output, try OpenAI, Anthropic, or Gemini's larger models.