After signup you receive a one time grant of 1,000 tokens and 100 MB of cloud storage. The free tier is for trying the app only; the provider does not guarantee ongoing availability.

You can cancel anytime in the subscription settings or by email, effective at the end of the current period. Without timely cancellation your subscription renews automatically. After cancellation you are downgraded to the free tier; fees already paid are not refunded.

You can upgrade at any time and it takes effect immediately. You are charged the pro rated price difference for the remaining billing period, and you receive the additional tokens immediately. Upgrade here: Subscribe

A downgrade only takes effect at the end of your current term. Until then you keep all features of the higher plan. Downgrade here: Subscribe

Each AI request consumes tokens for your input and the system context. We target about USD 1.00 per 100,000 tokens; at least one token is always used. The API and Gemini price lists help you estimate costs. Gemini pricing - OpenAI pricing
Audio transcription is billed per minute, not by prompt length. We speed up transcriptions to reduce cost but add approx. 20 percent tokens to the minute total. You can adjust the speedup factor in settings.

Unused tokens expire at the end of each billing month and cannot be carried over.

Packs remain on your account but are frozen. You can use them again when you upgrade back to Plus or Premium.

Backup lets you upload your files and app assets to the cloud to sync data across devices. Encryption happens only on your device using your cloud password; the provider never sees it. Backups are created automatically when data changes or when you trigger them manually. If you lose your cloud password you can reset your storage and set a new password by clicking the reset link in the confirmation email.

If after a plan change you exceed your new storage limit, the excess data will be kept for 30 days; it may be deleted after that. If you fully delete your account all cloud data can be retrieved for 30 days and is then permanently removed.

We are continually expanding our model lineup. Currently available: Gemini 2.5 Flash, Gemini 2.5 Flash Lite, Gemini Flash Pro, GPT-5 nano, GPT-5 mini, and GPT-5. We also support Whisper speech to text and vector text embeddings. Availability may vary by API or provider.

Switch to a more capable model. For critical tasks create a question and action pair to provide clearer context. Prompting in English often yields better results.

Partially. The app requires an internet connection for login and for all AI and cloud backup features. Only local scripts, assets, and saved chats can be opened without a connection.

By default recording stops when prolonged silence is detected. In settings you can increase the silence time so recording practically never ends until you click the microphone again.