← Docs
Getting Started

Bring your own API key

Create keys at Anthropic, OpenAI, Gemini, or run Ollama locally. Keys live in your keychain.

IntelliPrompter uses your own API keys for every AI call it makes. You create keys at the provider you want to use (Anthropic, OpenAI, Google, or Ollama for local), paste them into Settings, and the app routes work to those providers from then on. This page walks through why, which providers are supported, and how to set each one up.

Why BYOK

The short version

  • Your meetings are sensitive. Your prospect names, deal terms, internal strategy, and compensation talks live in your transcripts. Those should go to a provider you picked, under an account you control.
  • Your bill is yours. You pay the provider directly at wholesale rates. We do not resell inference or take a cut of your tokens.
  • No middle party sits in your API traffic. Calls go from your Mac to the provider you configured. We never see the request or the response.
Supported

Providers

Pick one or several. The app works with any combination.

  • Anthropic Claude. Highest coaching quality in our testing.
  • OpenAI. Solid all-rounder.
  • Google Gemini. Generous free tier for a long time.
  • Ollama. Fully local, free, private. Needs a model running on your Mac.
Anthropic

Create a Claude key

  1. Go to console.anthropic.com and sign in.
  2. Open API Keys and click Create Key.
  3. Minimum scope is a workspace key with default permissions. You do not need admin.
  4. Add billing under Plans & Billing. New accounts sometimes get free credits.
  5. Copy the key. You only see it once.
OpenAI

Create an OpenAI key

  1. Go to platform.openai.com and sign in.
  2. Open API Keys and click Create new secret key.
  3. A project-scoped key is fine. Default permissions work.
  4. Under Billing → Limits, set a soft spend limit of $10 to start. You can raise it later. This is the single best habit for first-time API users.
  5. Copy the key. You only see it once.
Google

Create a Gemini key

  1. Go to aistudio.google.com and sign in with Google.
  2. Click Get API key, then Create API key.
  3. Pick an existing Google Cloud project or let it make one for you.
  4. Gemini has a free tier that handles light to moderate use at no cost. You still get rate limits, but you do not need to add billing to start.
Ollama

Set up local inference

  1. Download Ollama from ollama.com and install it.
  2. Open Terminal and pull a model. A good default is ollama pull llama3.1:8b.
  3. Make sure the Ollama server is running. The menu-bar icon shows a steady llama when it is up.
  4. No API key is required. The app talks to Ollama over your local network.
NoteOllama runs the model on your Mac. Apple Silicon with 16 GB of RAM runs 8B models comfortably. Smaller machines should pick a smaller model like llama3.2:3b.
Storage

Where your key lives

Keys are stored in your Mac's Keychain. You can inspect them in Keychain Access under the service name IntelliPrompter. We never see them, never transmit them to our servers, and never back them up. If you wipe your machine, your keys go with it and you paste them in again.

Configure

Add a provider in the app

  1. Open Settings → AI Providers.
  2. Click Configure next to the provider you want.
  3. Paste the key you created. For Ollama, confirm the server is running.
  4. Click Save.
  5. The status flips to Connected. You are done.
Routing

Using several at once

You can add multiple providers. The app routes each task to the cheapest capable model without you touching anything. High-volume background work goes to a cheap or local model. The harder moments (long-form reasoning, final response generation) go to a stronger cloud model. You do not have to configure the routing. It just happens.

If you only configure Ollama, everything runs locally and your bill is zero. If you add a cloud provider on top, routing uses local for the volume tasks and cloud for the hard moments.

Cost

What to expect per month

  • Local-first user. Ollama handling volume plus a cloud key for the hard moments typically costs a couple of dollars a month for moderate use.
  • Cloud-only user. No Ollama, cloud for everything, moderate use. Expect $4 to $10 per month.
  • Heavy user. Many hours of meetings per week, cloud-only, no local. Can run higher. Local inference knocks this down fast.

For comparison, incumbents charge a flat $30 per month regardless of use. BYOK users holding their own key almost always come out ahead.

Revoke

Removing a key

You have two independent ways to cut off a provider:

  • Delete the key in the provider's console (Anthropic, OpenAI, or Google). The key stops working immediately for every app using it, including ours.
  • Remove the provider in Settings → AI Providers. The app forgets the key and stops routing work to that provider. The key itself remains valid in the provider's console until you delete it there.

We do not need to be in the loop either way. There is no “deactivate my IntelliPrompter account to stop API charges” step. The key is yours, the billing is yours, and revoking is one click at the source.

What's next

Where to go from here

More docs landing as the public beta gets closer.

Join the beta waitlist