When setting up a Bring-Your-Own-Key (BYOK) transcription workflow, most users immediately think of OpenAI's Whisper API. While Whisper is highly accurate and costs about $0.36 per hour, there is a much cheaper alternative that most people miss: Google Gemini 1.5 Flash.
Unlike older models, Gemini 1.5 Flash accepts raw audio files directly as native inputs (using 32 tokens per second of audio). Combined with Google's aggressive pricing on AI Studio ($0.075 per 1 million input tokens), transcribing and summarizing a one-hour meeting costs approximately $0.015.
This guide explains how to get a Google Gemini API key and configure it for ultra-cheap transcription.
The Economics: Gemini vs. Whisper
| Model | Cost per 1 Hour | Cost for 100 Hours/Month | Billing Model |
|---|---|---|---|
| OpenAI Whisper API | $0.36 | $36.00 | Pay-as-you-go |
| Google Gemini 1.5 Flash | ~$0.015 | $1.50 | Pay-as-you-go (Free tier available) |
| SaaS Meeting Subscriptions | N/A | $15.00 - $20.00 | Flat monthly rate |
[!TIP] Google AI Studio provides a free tier for Gemini 1.5 Flash that allows up to 15 requests per minute and 1,500 requests per day. For personal note-taking, you can transcribe your meetings entirely for free.
Step-by-Step: How to Get a Google Gemini API Key
To use Gemini for transcription, you need to generate an API key from Google AI Studio (formerly MakerSuite):
Step 1: Go to Google AI Studio
- Open your web browser and navigate to Google AI Studio.
- Log in using your standard Google account (or Workspace account).
Step 2: Create the API Key
- In the left-hand navigation panel, click the Get API Key button.
- Click Create API Key.
- Select whether you want to associate the key with a new Google Cloud project or an existing one (creating a new project is simplest).
- Google will generate a string starting with
AIzaSy.... Copy this key immediately.
Step 3: Enable Billing (Optional for higher limits)
If you run into rate limits on the free tier, you can transition your Google Cloud project to the Pay-As-You-Go plan. You will only be billed for what you actually use, which will be pennies per month for meeting notes.
Configuring the Key in justREC
Once you have your Gemini API key, integrating it into your meeting workflow takes seconds:
- Open justREC on your Mac.
- Click the settings icon in the app window or open preferences.
- Under the AI Providers list, select Google Gemini.
- Paste your copied API key into the input field.
- Save the configuration.
Now, whenever you finish recording a meeting or import an audio file, you can choose Gemini 1.5 Flash as your primary transcription and summarization engine.
Why Gemini 1.5 Flash is Ideal for Meetings
- Massive Context Window: Gemini 1.5 Flash has a context window of 1 million tokens. This means you can feed it a 5-hour recording, and it will process the entire audio stream and prompt history in a single request.
- Multimodal Native Audio: Because the model processes the raw audio directly (rather than transcribing it to text first and then running it through a text LLM), it can capture tone, pauses, and context more effectively.
- Speed: Gemini 1.5 Flash is exceptionally fast. A one-hour meeting is typically transcribed and summarized in under 30 seconds.
Start Saving on Meeting AI
Stop paying markup on subscription-based meeting notes. Download justREC for macOS, connect your Google Gemini API key, and enjoy professional-grade meeting transcripts and summaries for less than two cents per hour.