Answers to the questions learners ask most

SpeakShark is an AI-powered English speaking practice platform with four native-accent 3D AI teachers, real-time pronunciation scoring, and structured speaking modes. It's designed for learners who can already read and write English but freeze in real spoken conversations.

Yes. The free tier requires no credit card and includes 3 sessions per day with one AI teacher (Sarah), 2 topic categories, and a 7-day session history. You'll never be charged unless you choose to upgrade to Pro.

Any modern browser with a microphone — phones, laptops, tablets, desktops. There is no separate app to install. The web app works on iOS, Android, Windows, and macOS browsers.

No. The microphone built into your phone, laptop, or any standard headset works. A reasonably quiet environment helps the speech recognition pick up your voice clearly.

No. Speech recognition and AI conversation require a server connection. A typical home or office Wi-Fi connection is sufficient.

SpeakShark teaches English only. The user interface itself is available in 19 languages including English, Vietnamese, Japanese, Korean, Chinese (Simplified), Indonesian, Thai, Spanish, Portuguese, Hindi, Turkish, French, German, Arabic, Russian, and Italian.

Pronunciation scoring uses a production-grade speech recognition engine for transcription combined with a separate phoneme-level scoring pass. The engine handles accented English well, including non-native pronunciation. Scores are intended as a directional signal to identify which sounds are causing problems — not as a substitute for a human evaluator.

A production-grade multilingual ASR engine running on a low-latency inference platform. The engine was trained on hundreds of thousands of hours of supervised speech data and is robust to accents, background noise, and non-native pronunciation. Specific vendor and model are not disclosed for operational reasons.

A purpose-tuned conversational language model, with a per-teacher system prompt encoding persona, accent target, CEFR difficulty band, and pedagogical strategy (gentle correction, scaffolded follow-ups, vocabulary expansion). Specific vendor and model are not disclosed.

Pronunciation scoring is grounded in deterministic acoustic analysis of your actual speech, which limits hallucination on the scoring side. The conversational reply is AI-generated and can occasionally produce a phrasing that is technically correct but unusual. Where pronunciation matters, compare your audio against the native sample provided.

Internal data suggests learners who practice 10+ minutes daily for 30 days improve their overall speaking score by an average of 15 points. Individual results vary widely based on starting level and consistency. The platform is built around the assumption that consistency beats intensity for spoken language.

For everyday speaking practice, AI conversation produces meaningful improvement at low cost. For high-stakes coaching with a specific human in mind, human teachers still add value AI does not yet match. Most learners benefit from using both.

ELSA emphasizes phoneme drill exercises with mouth-position graphics. SpeakShark emphasizes open spoken conversation with phoneme-level corrections inside the conversation. ELSA is better for intensive isolated drill; SpeakShark is better for free-form conversational practice with real-time feedback.

• Full SpeakShark vs ELSA Speak comparison

Cambly uses human tutors and charges per minute. SpeakShark uses AI teachers and offers a real free tier. For consistent daily speaking practice without per-minute cost, SpeakShark wins on price and 24/7 availability. For nuanced human feedback on specific high-stakes situations, Cambly retains advantages.

• Full SpeakShark vs Cambly comparison

ChatGPT Voice is a general assistant. It does not score pronunciation, does not commit to a target accent, does not track progress, and costs more per month than SpeakShark Pro. If your goal is to actively improve your speaking — which requires measurement and structured feedback — the dedicated tool gives you what ChatGPT Voice doesn't.

• Full SpeakShark vs ChatGPT Voice comparison
• SpeakShark vs ChatGPT (text mode)

Duolingo is a multi-language general app with light AI speaking features in its premium tier. SpeakShark is single-purpose English speaking with deep per-utterance feedback. Use Duolingo for vocabulary and grammar foundations across languages, SpeakShark for English speaking-specific deep work.

• Full SpeakShark vs Duolingo comparison

No. SpeakShark teaches you to speak intelligibly and improve over time in a target accent of your choosing — but the goal is communicative fluency, not erasing how you naturally sound. If accent reduction specifically is your goal, dedicated accent-coaching products may fit better.

• SpeakShark vs BoldVoice (accent coaching)

Audio is processed for transcription and scoring. Sessions are recorded so you can listen back during the session, and Pro users can review past session recordings as part of full history. Audio is never used for advertising or sold to third parties.

Profile data, session history, and stored audio references are deleted. Aggregated, non-identifiable analytics may be retained for product improvement.

User accounts and session metadata are stored on managed PostgreSQL infrastructure. Audio is processed by a production-grade ASR engine for transcription. We do not transfer your data outside these processors for marketing purposes.

You can request data export and deletion via support at speaksharksupport@gmail.com at any time. Standard data-handling practices are documented in the privacy policy at /privacy.

You can wait until the next day for the free quota to refresh, or upgrade to Pro for unlimited sessions. There's no credit card requirement on the free tier, so you'll never be charged unexpectedly.

Pro is $12 per month, or $100 per year (~$8.33/month, save 30%). It unlocks unlimited sessions, all 4 AI teachers, all 50+ topic categories, Speaking Challenges, Role Plays, full session history, detailed analytics, and phoneme-level error breakdown.

Cancel from the billing page in your account settings. The subscription remains active until the end of the current billing period, then reverts to the free tier — no further charges.

Discount eligibility is reviewed on request via support. SpeakShark was built by students and is sympathetic to student budgets — write to speaksharksupport@gmail.com with proof of enrollment.

Refunds within 7 days of subscribing are typically honored if you have not used Pro features extensively. Contact support to request a refund.

The browser must have microphone permission for speakshark.com. Check the browser's address bar microphone icon, or your operating system's privacy settings, to grant permission. Reload the page after granting access.

Common causes: background noise (try a quieter room), microphone too far from your mouth (try moving closer or using a headset), speaking too quickly without natural breaks, or mismatch between your target accent and the selected teacher's accent.

VRM avatars require WebGL. Try a different browser (latest Chrome, Edge, or Firefox), confirm hardware acceleration is enabled in browser settings, or update your graphics drivers.

AI conversations are mostly good but occasionally drift. Restart the session — each session is independent, so you can start fresh without losing earlier progress. If a specific issue keeps happening, send the session details to support.