Last updated: September 28, 2026

Gemini API Key Setup Guide.

Trace uses your own Gemini API key. Create a key in Google AI Studio, then paste it into Settings in the app.

Privacy summary:Trace does not include a built-in Gemini account or shared API key. To transcribe, organize, and analyze meetings, audio, transcripts, meeting background, and generated results are sent to the Gemini API. For details, see the Privacy Policy.

Create a Gemini API key

  1. Open Google AI Studio. Go to Google AI Studio and sign in with your Google account.
  2. Open the API key page. Find the API key area in AI Studio, or use Google's Gemini API key documentation.
    Google AI Studio API Keys page showing creation, grouping, filtering, and column controls; account, key, project identifier, and usage data are outside the crop
    The API Keys page provides controls to create a key, group by key or project, and filter by project.
  3. Create a new key. Follow the Google AI Studio prompts to create a key. Treat it like a password, and do not paste it into public documents, screenshots, chat messages, or support reports.
    Google AI Studio Create API key dialog showing the key-name field and project-selection prompt; the actual project is outside the crop
    The Create API key dialog asks you to choose a project. After creating the key, store it like a password and do not paste the full key into chats, documents, or support reports.
  4. Paste it into Trace. Open Settings in Trace, paste the key into the Gemini field under API Keys, then tap Add Key. The key is stored only in iOS Keychain.
    The API Keys section of Trace Settings: the Gemini field has no key stored yet, with Label and API key fields and an Add Key button
    The API Keys section in Trace. Keys in the same project share one allowance; when one key reaches its quota, Trace can switch to another available key.

Free API keys and free limits

Google AI Studio can usually create a free API key, and some models may include free-tier or trial limits. Free quotas, daily request counts, requests per minute, regions, and model availability can change. Always use Google's official pricing and limit pages as the source of truth.

Trace's available models follow its public model catalog. Free quota varies by account, project, model, region, and Google policy, so Trace cannot promise a fixed number of meeting hours. Limits apply per project, so several keys in the same project share one allowance, and requests per day reset at midnight Pacific time. Start with a short recording to confirm that the key and quota work before processing longer audio. See Trace Models & Costs for paid estimates, and use the Google AI Studio Rate Limit page for current limits. When Gemini returns a rate limit, Trace waits before retrying.

How free-tier data is used: Google may use content sent on the free tier to improve its products, human reviewers may process it, and Google asks you not to send sensitive, confidential, or personal information on free services. For those meetings, use a project with billing enabled. See free API key essentials and the Privacy Policy.

Google AI Studio Rate Limit page showing model, category, RPM, TPM, and RPD columns; project, billing, and actual usage data are outside the crop
The Rate Limit table shows RPM, TPM, and RPD limits by model. Actual limits vary by account, project, model, and Google policy.

Paid API keys and paid limits

If the free tier is not enough, or if you want to process longer audio more reliably, enable billing for the Google project that owns the API key. Trace usage is billed by Google according to your project, model, region, limits, and current Gemini API pricing. Long-audio cost mainly depends on audio input tokens, transcript output tokens, and meeting-note analysis input / output tokens. For the full formula and examples, see Trace Models & Costs. Trace keeps pricing transparent and does not add a backend processing fee, so you can use Google's public billing standards directly.

Trace design

Bring your own API key

Trace does not host a Gemini account or include a shared API key. Usage, pricing, and limits are controlled by your own Google project.

Local meeting records

Meeting notes, transcripts, generated results, and processing state are stored on your device so you can review, retry, and share them later.

Long audio import

Trace is designed around recorded or imported meeting audio and helps turn full meetings into reusable records.

One-time unlock

The app uses a one-time unlock model. Gemini API usage is charged by Google based on your API key's project and billing setup.

Meeting background

When you record or import audio, Trace lets you add meeting background such as customer names, product names, agenda topics, acronyms, and speaker context. This text is sent to the Google Gemini API together with the audio, transcript, template, and generated draft so Gemini can recognize domain-specific terms more accurately.

Keep the background short and factual: one to five bullets is enough for most meetings. Do not paste API keys, passwords, private credentials, or information you do not want processed by Gemini.

Meeting background adds a small amount of text input tokens. For the cost model, see Models & Costs.

Data sent to Gemini

To complete transcription, meeting-note generation, analysis, and fact-checking, Trace sends imported audio, transcripts, meeting background, and generated or app-organized results to the Google Gemini API.

Your Gemini API key is stored only in iOS Keychain. It is not written into meeting notes, transcripts, Markdown exports, or in-app meeting data. For more about data handling, see the Trace Privacy Policy.

Official documentation