Configure AI providers, speech, translation, and learning preferences
Theme
Follows your operating system preference
Interface Language
Choose the language used across the EchoType interface. This does not change your learning target.
Learning Target
The language you are learning
Provider
Groq
DefaultLlama, Gemma via LPU inference
Custom API endpoint. Use a proxy URL to route through a third-party service.
Default: /openai/v1/chat/completions. Change this only if your proxy uses a different path.
Model
Fast & capable
Max Output Tokens
4096 tokensUsing global setting (4096 tokens)
Max Output Tokens
4096 tokensLimits the maximum number of tokens the AI can generate per response. Lower values save costs and speed up responses. Higher values allow longer, more detailed answers.
Voice source
Speed
1.0xPitch
1.0Volume
100%Target language
Listen
On NowShow translations while listening to content and sentence highlights.
Read
On NowShow translations while reading aloud and reviewing the transcript.
Speak
Off NowShow translations while practicing speech recognition feedback.
Write
On NowShow translations while typing practice content and checking answers.
Listen, Read, and Write start with translation on. Speak starts with translation off. Each module keeps its own current visibility state.
Enable selection translation
Select text to translate and add to favorites
Enable smart auto-save
Automatically collect words based on practice patterns
Sensitivity
Daily cap
Enable recommendations
Show AI suggestions at the bottom of practice pages
Number of results
Enable Shadow Reading
Link content across Listen, Read, and Write modules
Shadow reading is a technique where you practice the same material across skills: Listen β Read aloud β Write. When enabled, switching between modules highlights your current content so you can continue seamlessly.
For the Speak module, related conversation scenarios are recommended based on your content topic.
Assessment Provider
SpeechSuper API
Get phoneme-level pronunciation scoring. The free tier includes 1,000 calls per month.
App Key
Secret Key
Monthly Limit
Pronunciation assessment runs after each speech input in the Speak module. SpeechSuper gives you phoneme-level IPA analysis, while the AI fallback provides estimated scores and tips based on text comparison.
Assess your English level
Take a quick test (CEFR A1-C2) to get personalized learning recommendations.
Keyboard shortcuts
Search opens with Ctrl+K by default on this device. Open the editor to customize bindings and check which scopes are active on the current page.
Download all content items as JSON
Download records and sessions as JSON
Download all data including favorites, conversations, books, and settings
Manage preset tags for quick selection when importing content. Maximum 20 tags.