🎙️ AI-Powered Full Duplex Voice

Communicative English AI Coach

Practice spoken English with a real-time AI coach. 8 practice modes from casual conversation to IELTS prep. Get instant feedback on fluency, grammar, and pronunciation.

💬
Free Conversation
Open-ended AI conversation on any topic. AI adapts to your level. Build confidence naturally.
All Levels⭐ Popular
🏢
Workplace English
Professional scenarios — meetings, phone calls, giving feedback, presentations, team standups.
Intermediate
📝
IELTS / TOEFL Prep
Timed speaking test simulation. Part 1, 2, 3 with IELTS band scoring and detailed feedback.
Advanced⭐ Popular
🔤
Pronunciation Drills
AI says a word or sentence, you repeat. Get scored on accuracy. Targets common challenges.
Beginner
📖
Grammar Coach
AI chats with you and pauses to correct grammar mistakes in real-time. Tracks error patterns.
Intermediate
🗣️
Debate & Discussion
AI takes opposing stance on a topic. Practice structured arguments, rebuttals, and persuasion.
Advanced
🎭
Role-Play Scenarios
Real-world situations — restaurant, airport, doctor visit, landlord negotiation, complaints.
All Levels
📊
Presentation Practice
Deliver a 2-5 min presentation. AI scores structure, clarity, filler words, pacing, and confidence.
Advanced

📚 8-Week Learning Path

💬

Free Conversation

Configure your session and start practicing
⚠️
🔑
Gemini API Key (Required)
Full-duplex voice powered by Google Gemini Live API
Key stored in browser session only. Get free Gemini API Key →
CONNECTING
5:00
Connecting to AI Coach...

Session Complete 🎉

Here's your performance breakdown

Estimated CEFR Level
B1

Frequently Asked Questions

What is the Communicative English AI Coach? +
The Communicative English AI Coach is a free, AI-powered spoken English practice platform by Simpatico HR. It uses Google Gemini's Live API for full-duplex (real-time two-way) voice conversation — meaning you speak naturally and the AI responds instantly, just like a human tutor. It offers 8 practice modes, real-time scoring across 5 metrics, CEFR level estimation (A1–C2), downloadable certificates, and a structured 8-week learning path.
Is this free? What do I need to use it? +
Yes, it is completely free. You need: (1) a Simpatico HR Career Lab account, (2) a Google Gemini API key (also free from aistudio.google.com/apikey), and (3) Chrome or Edge browser for microphone + speech recognition support. There are no subscriptions, usage limits, or hidden fees.
How is this different from ELSA Speak, Cambly, or Duolingo? +
vs ELSA Speak ($50/yr): ELSA uses record-and-score; we use full-duplex real-time voice. vs Cambly ($100+/yr): Cambly uses human tutors; we use AI available 24/7 for free. vs Duolingo: Duolingo is text/tap-based; we offer real spoken conversation practice. Our unique advantage: full-duplex AI voice + 8 specialized modes + live scoring + gamification — all free.
Can I use this for IELTS Speaking test preparation? +
Yes! The IELTS Speaking Test mode simulates all 3 parts of the real IELTS exam — Part 1 (Introduction), Part 2 (Cue Card), and Part 3 (Discussion). The AI acts as a certified IELTS examiner, asks realistic questions, and provides estimated band scores. You can practice unlimited times with no cost.
What are the 8 practice modes? +
(1) Free Conversation — Open chat on any topic. (2) Workplace English — Meetings, interviews, client calls. (3) IELTS Speaking Test — Full exam simulation. (4) Pronunciation Drills — Difficult sounds, minimal pairs. (5) Grammar Coach — Real-time grammar corrections. (6) Debate & Discussion — AI takes the opposing stance. (7) Role-Play — Restaurant, airport, doctor scenarios. (8) Presentation Practice — Structure, clarity, filler scoring.
How does the scoring work? +
You are scored in real-time across 5 metrics: Fluency (words per minute, pause patterns), Grammar (error-based deductions from AI corrections), Pronunciation (browser SpeechRecognition confidence scores), Vocabulary (unique word ratio + lexical diversity), and Confidence (filler word frequency). After each session, you receive a CEFR level estimate (A1–C2) and can download a certificate.
What browsers are supported? +
Chrome and Microsoft Edge on desktop/laptop are recommended. The full-duplex voice engine works on any modern browser with WebSocket support, but the SpeechRecognition API (used for pronunciation scoring and transcript generation) is currently only available in Chromium-based browsers.