dot.dictate

dot.dictate · Android

Help & setup

From installation to your first dictation. Choose a topic or go straight to the message shown in the app.

Your first dictation

  1. Open dot.dictate, choose your AI provider, paste your own API key and save it. Use the guide below if you do not have a key yet.
  2. Allow microphone access. Enable dot.dictate in Android’s keyboard list. The setup actions in the app open the relevant settings.
  3. On Samsung, enable “Keyboard button on navigation bar”. If Samsung’s voice-input shortcut takes over, set Samsung Keyboard → Voice input to “None” and use the keyboard selector instead. Menu names vary by One UI version.
  4. Choose the spoken language and output language. “Automatic” detects the spoken language; “Same as input” keeps it rather than translating.
  5. Open a text field in any app, use the keyboard selector and choose dot.dictate. With automatic start enabled, recording starts when the microphone is ready. Otherwise, tap the record button.
  6. Speak, then tap “Finish and insert”. Review the result before sending it. The app returns to your normal keyboard after insertion.

Get your API key

OpenAI

  1. Sign in to OpenAI Platform. In API keys, create a new secret key for a project dedicated to dictation.
  2. Copy the secret when it is created, then select OpenAI in dot.dictate, paste it into “API key” and save. Do not use an organization Admin API key.
  3. In Billing, check whether API billing is enabled and you have available credit. Set spending limits and alerts to suit your budget.
  4. Tap “Test” in dot.dictate. A ChatGPT subscription does not by itself pay for your API usage.

Google Gemini

  1. Sign in to Google AI Studio with your Google account and open API keys.
  2. Create a key for a project you control, using the current key type offered by AI Studio. Avoid old, blocked or unrestricted keys.
  3. Select Gemini in dot.dictate, paste the key and save it. Tap “Test” to check access to the models used by the app.
  4. Check the project’s API plan, quota and billing in AI Studio. Availability and free or paid usage depend on the model, country and account; a Google AI subscription is not a guarantee of API access.

xAI Grok

  1. Sign in to the xAI console and select or create the team that will own the API key.
  2. Open API Keys and create a key with access to the speech and text models required by the app.
  3. Check credits and billing in the console. A consumer Grok subscription does not guarantee that your API team has available credit.
  4. Select Grok in dot.dictate, paste the key, save and tap “Test”. If model access is denied, check the key’s permissions and the team’s model availability.

Costs & credits

dot.dictate and the AI provider are separate services. App trial or licence access does not include API credit. Audio transcription, refinement, translation and automatic UI translation may consume provider usage. Prices, quotas and model availability are set by the provider; check its console before adding credit. The app does not display a verified account balance.

Problems & solutions

“Recording kept” means you can retry the pending recording. It is not a permanent archive: do not cancel it if you still need that audio.

No API credit · recording kept

The provider explicitly reports exhausted credit. Check the account or team used by your key, add credit in Billing, then return and retry. This is not an internet failure.

API quota unavailable · recording kept

The response does not establish that your balance is zero. Check billing activation, available credit, project quotas and the account associated with the key. Do not add money blindly.

Billing limit reached · recording kept

A billing cap was reached. Check Billing and the applicable limits. If you choose to raise a cap, review your budget first; adding credit alone may not remove the cap.

API limit reached · recording kept

An organization or project spending/usage limit was reached. Check the limit shown by your provider. Raise an allowed spending limit or wait for its reset; a provider-approved usage limit may require a request for an increase.

Too many requests

Wait briefly, then retry. This usually concerns request or token throughput, not your balance. Repeated attempts can prolong the problem.

Invalid API key / API access denied

Check that the provider matches the key, that it has not expired or been revoked, and that the required models are allowed. Replace a compromised key in the provider console and in the app.

Not sent / internet or timeout error

Check Wi-Fi or mobile data and any VPN. For a transient provider outage, wait and retry. If the app says “recording kept”, retry before cancelling; cancelling discards that pending recording.

The AI provider returned no text.

The response contains no transcription; this does not prove a credit problem. Make sure microphone access is allowed, speak after recording begins, and try a short sentence. Quiet or empty audio does not require extra credit.

Provider error / model unavailable

Try again later for a temporary outage. If a model is unavailable or access is denied, check the provider and key permissions. Contact support with the app version and error code, not your key or dictated content.

Your data

Your key is encrypted on your phone using Android Keystore. Dictation audio goes to the selected AI provider over HTTPS; text may also be sent for refinement or translation. Providers apply their own data policies. Do not share API keys, and never send them to support. The detailed policy explains diagnostics, optional statistics and the licensing service.

Read the privacy policy →

Still stuck?

Send the app version, selected provider and exact error message to contact@dotdesign.ro. Never include a secret key, audio or private conversation.

contact@dotdesign.ro