Aloud
Aloud uses your own API keys. Nothing you capture leaves your device except the requests you trigger, and you pay providers directly rather than paying a subscription.
You need two keys to start, and a third if you want to dictate:
| Key | What it does | Where to get it |
|---|---|---|
| Anthropic | Translates and groups your sentences | console.anthropic.com |
| Google Cloud | Generates the audio | console.cloud.google.com — enable the Cloud Text-to-Speech API, then create an API key |
| OpenAI (optional) | Transcribes dictation | platform.openai.com |
Add them under Bank → Settings → Keys. They are stored in the iOS Keychain.
Tap the mic in the middle of the tab bar and either dictate or type sentences you would actually say.
On the Today screen, tap Prepare them. Aloud translates your sentences, groups them into topics, and generates audio. This happens once per sentence per language.
Play your sentences on a loop and repeat out loud over the top. Three modes: bilingual, target-only, and gapped, which leaves silence for you to answer into.
Say each sentence from memory before checking. Missed sentences return before the session ends.
Sentences are either shared across every language or pinned to one. Open the Bank, tap a sentence to expand it, and you'll see which — tap to change it. Shared sentences still need Prepare them run once in each language.
Check that the Cloud Text-to-Speech API is enabled on the same Google Cloud project your API key belongs to, and that billing is enabled on that project.
iOS stretches audio to slow it down without changing pitch, and that processing degrades below roughly 0.5×. Try 0.6–0.75× instead.
Check usage in each provider's console. Translation is billed per token and is small. Speech is billed per character, generated once per sentence, then cached on your device permanently.
No. Everything is stored on-device with no backup service, so deleting the app deletes your bank.
davidmacdonald41042@gmail.com — this is a personal project, so replies may take a few days.