Connection and API key guide
To connect Prompt Paul to your own model, use an OpenAI-compatible HTTP(S) endpoint or the official Anthropic API format on Anthropic's API domain. The services below are common examples, but Prompt Paul does not bundle provider accounts and compatibility, available models, authentication, and pricing vary by service. LM Studio can expose a local OpenAI-compatible endpoint; the extension still requires a non-empty API key field for API profiles, even when the local server does not use authentication. Codex and Hermes use their own sign-in or gateway credentials. This guide covers common starting points and key safety.
Some services may offer free tiers or promotional credits, but availability and limits change. Check each provider's current terms. LM Studio runs models locally, so it avoids provider per-request charges, although your computer still uses its own resources. API profiles still require a non-empty API key field in the extension.
An API key is a unique identifier that authenticates your requests to a service's API. Think of it as a password that allows Prompt Paul to communicate with the provider's servers. Codex subscription profiles use Sign in with ChatGPT instead of an API key, while Hermes has its own optional gateway authentication.
Prompt Paul can use Brave's Search API for model-directed web grounding. Create a key from the Brave Search API page, then save it in the extension's Web Search Grounding section. Enable Web in the side panel only for runs where live sources are useful.
Search is off by default, limited to one search attempt per run, and may be skipped by the model. Brave pricing, credits, and quotas depend on its current plan. The key stays in Chrome storage and is never sent to your model provider or analytics.
Open the OpenAI API keys page and sign in or create an account if needed.
On the API keys page, choose Create secret key (the exact label may change), then copy the key and store it somewhere safe. You will not be able to see the full secret again.
OpenAI may require billing information before using their API.
Always keep your API keys secure and do not share them.
Open the Anthropic API keys page and sign in or create an account. Create a key, copy it once, and store it securely.
Use the official Anthropic API endpoint, your key, and the model name in an API profile. Check Anthropic's current documentation for model availability, media support, and pricing.
Visit the Google AI Studio website and sign in with your Google account. Use the key with a Prompt Paul-compatible endpoint; a native Gemini API endpoint is not automatically compatible with Prompt Paul.
After logging in, look for the "Get API key" button, usually located in the top right corner. Click "Create API key in new project".
You will be prompted to create a new Google Cloud project or use an existing one. Follow the instructions. Once the project is set up, the API key will be generated and displayed. Copy and securely store it.
Using Google AI APIs may incur costs. Review Google's pricing policies before making API calls.
Google Cloud Console might be needed to manage projects and API settings.
Go to the Groq website and create an account.
Navigate to the API section of your account after logging in. You should see an option to generate a new API key. Copy and store the generated API key securely.
Groq's API access might be subject to usage quotas and pricing. Check their documentation for the latest details.
Go to OpenRouter and sign up for an account.
After logging in, navigate to your account settings or API keys section. Click "Create API key" or a similar button. Copy the generated key and keep it safe.
OpenRouter provides access to multiple LLMs through a single API. Availability, pricing, free-model access, and billing requirements depend on the selected model and current OpenRouter terms.
Go to the Together AI platform and create an account.
Locate the API key section within your account after logging in. Click the "Create API Key" button. Copy the generated API key and save it securely.
Together AI also offers billing plans. Check their website for details.
LM Studio is a local application, not a cloud service. It can expose an OpenAI-compatible endpoint for models running on your computer. Prompt Paul's API profile still requires a non-empty API key field, even if the local server ignores authentication.
Download and install LM Studio from its official website. After installation, you can download various models and use them locally.
Because the models run locally, performance depends on the hardware of your machine.
This guide provides a general overview. Always refer to the specific documentation and instructions provided by each platform for the most accurate and up-to-date information.
