தமிழறிவு Thamizharivu
🌿 Open-source · Tamil-first · Kid-friendly

தமிழ் — Living,
Breathing,
Growing

உச்சரிப்பு சரி பார்க்க · எழுத்து திருத்த · இலக்கணம் கற்க

An open AI platform that listens to your Tamil, understands every accent — from Chennai streets to Sri Lankan shores — and gently guides children back to their mother tongue.

யாதும் ஊரே யாவரும் கேளிர்
— புறநானூறு · Purananuru 192  |  "Every land is my home; all people are my kin."
80M+
தமிழ் பேசுபவர்கள்
Tamil Speakers
2,000+
ஆண்டு பழமை
Years of Literature
5+
வட்டார வழக்குகள்
Regional Accents
1,330
திருக்குறள் வெண்பாக்கள்
Thirukkural Couplets

தாய்மொழியை மீட்போம்

Tamil is one of the world's oldest living classical languages — yet millions of children today grow up unable to read, write, or speak it fluently. English-medium schooling, migration, and digital gaps have eroded what 2,000 years of literature built.

Thamizharivu is our answer. We build open AI tools that listen to Tamil spoken in every accent, correct gently without shaming, and celebrate the diversity of our dialects — from the fast rhythm of Chennai to the melodic cadence of Jaffna.

We do not impose a single "correct" Tamil. We honour every voice — Madurai, Coimbatore, Sri Lankan, Malaysian, diaspora — while helping children write beautifully and speak clearly.

🎙 கேட்டல் — Listening First

AI that hears the child's voice before it speaks. Accent-aware speech recognition trained on regional Tamil audio datasets.

✍️ திருத்துதல் — Gentle Correction

No red marks. No shame. Visual cues in Tamil guide the learner toward correct spelling and grammar without discouragement.

📖 வளர்த்தல் — Growing Together

Open datasets contributed by communities, annotated for accent and age, to train better models for every generation.

🌏 சேர்த்தல் — Inclusivity

Tamil belongs to all who speak it — in Tamil Nadu, Sri Lanka, Singapore, Malaysia, and the global diaspora.

கற்கக் கசடறக் கற்பவை

Learn to read, write, and speak Tamil — from the first stroke of அ to composing full sentences. All lessons are in Tamil-first, with English support available.

அகர வரிசை
Tamil Alphabet

Learn all 247 Tamil letters — 12 vowels (உயிர்), 18 consonants (மெய்), and 216 compound letters (உயிர்மெய்). Interactive stroke-order animations.

🎵
உச்சரிப்பு பயிற்சி
Pronunciation Practice

Speak a word aloud. Our AI compares your audio to reference phonemes and gives instant feedback — where your tongue should be, how long to hold the vowel.

📝
இலக்கணம்
Grammar Foundations

Tamil grammar — எழுத்திலக்கணம் (phonology), சொல்லிலக்கணம் (morphology), and பொருளிலக்கணம் (semantics) — explained simply with real examples.

📚
கதைகள் & பாடல்கள்
Stories & Songs

Public domain folk tales, Aesop in Tamil, and classic Sangam poems — read aloud, annotated, and searchable. Learn vocabulary through context.

✏️
எழுத்துப் பிழை திருத்தம்
Spelling Correction

Type or paste any Tamil text. Our fine-tuned language model highlights errors, explains the correct spelling, and shows the grammatical rule — in Tamil.

🏆
சாதனை பதக்கங்கள்
Gamified Progress

Earn stars for daily practice, unlock badges for mastering letter groups, and track your streak on a Tamil-language progress board. Learning feels like play.

ஒவ்வொரு குரலும் தமிழே

There is no single "pure" Tamil — there is the Tamil of your street, your district, your island, your diaspora. Our AI recognises and honours every variation, never forcing one accent as the only standard.

🏙️
சென்னை தமிழ்
Chennai Tamil

Urban rhythm, code-mixing with English, the baseline dialect for most media.

🌾
மதுரை தமிழ்
Madurai Tamil

Rich retroflex sounds, classical vocabulary preserved in everyday speech.

கோவை தமிழ்
Coimbatore Tamil

Softer consonants, distinct intonation, influenced by Kannada border regions.

🌊
ஈழத் தமிழ்
Sri Lankan Tamil

Preserves archaic grammatical forms lost in mainland Tamil; unique phonology.

🌴
மலேசியத் தமிழ்
Malaysian Tamil

Influenced by Malay and English, distinct musical quality, plantation vocabulary.

✈️
புலம்பெயர் தமிழ்
Diaspora Tamil

Children raised abroad blending Tamil with English, French, German, or Dutch.

எப்படி வேலை செய்கிறது?

From the moment a child speaks into the microphone to the feedback they see on screen — every step is open-source, Tamil-first, and privacy-respecting.

1
🎤
கேட்டல் · Listen

Child speaks a Tamil word. Audio captured at 16kHz mono.

2
🔊
அலசல் · Analyse

Whisper / Wav2Vec2 model transcribes and detects the accent region.

3
⚖️
ஒப்பீடு · Compare

Phoneme embeddings compared to native-speaker reference for that accent.

4
✍️
திருத்தம் · Correct

LLM checks spelling and grammar. Highlights error, suggests fix in Tamil.

5
🌟
பாராட்டு · Reward

Visual badge, star, or emoji reward reinforces correct pronunciation.

இப்போதே முயற்சி செய்க

Type a Tamil word or sentence. Get spelling and grammar feedback — powered by a Tamil-tuned language model.

எழுத்துப் பிழை திருத்தம் — Spelling Checker

ஒரு தமிழ் வாக்கியம் தட்டச்சு செய்யவும்

அவள் பாடம் படிக்கிறாள் நான் தண்ணீர் குடிகிறேன் அவன் வீட்டிற்கு வந்தான்

குழந்தைகள் தமிழில் விளையாடட்டும்!

Our kid-friendly interface is entirely in Tamil. Bright visuals, simple words, and a gamified experience that makes learning feel like a game — not a chore.

🔤
எழுத்து விளையாட்டு

Drag-and-drop Tamil letters to form words. Hear each letter pronounced correctly as you place it.

🎯
சொல் போட்டி

Flashcard challenge — see a picture, type the Tamil word. Three lives, five levels, unlimited fun.

🎙️
பேசு & கற்று

Speak a word shown on screen. The mic listens. A green ✅ or a gentle hint appears — no harsh buzzer.

📖
கதை வேளை

Illustrated Tamil folk tales read aloud by native voices. Tap any word to hear it again.

நட்சத்திர பரிசு

Earn stars, unlock badges, and climb the leaderboard — all described in Tamil, so reading is also rewarded.

🎨
எழுதும் பலகை

Draw Tamil letters with your finger or mouse. AI checks your stroke order and celebrates correct form.

காலத்தால் அழியாத தமிழ்

We begin with the Thirukkural — 1,330 couplets of timeless wisdom in the public domain. Every child learns Tamil through Valluvar's words.

திருக்குறள் · குறள் 391
கற்க கசடறக் கற்பவை கற்றபின்
நிற்க அதற்குத் தக.
பொருள்: கற்கத் தக்கவற்றை குற்றமறக் கற்றுக்கொண்டு, அதற்கேற்ப நடந்து கொள்க.

Meaning: Learn thoroughly what is worth learning; then live in accordance with what you have learned.

AI கட்டமைக்க தரவுகள் தேவை

Good Tamil AI needs good Tamil data. We collect, annotate, and release datasets under open licences. Every dataset starts with ethically sourced, community-contributed recordings.

🔤
எழுத்து தரவுத்தொகுப்பு
Letters Dataset

Handwritten and digital Tamil letters — all 247 characters across multiple fonts and handwriting styles. Useful for OCR training.

🎧
ஒலி தரவுத்தொகுப்பு
Audio Dataset

Native speaker recordings of Tamil phonemes, syllables, and common words — tagged by region, age group, and gender. CC BY 4.0.

📚
கதை தரவுத்தொகுப்பு
Stories Dataset

Tamil folk tales, Panchatantra, and simplified Sangam excerpts — sentence-aligned with audio, suitable for TTS and language model fine-tuning.

🗺️
வட்டார ஒலி தரவு
Accent Dataset

Region-tagged audio from Chennai, Madurai, Coimbatore, Sri Lanka, and Malaysia. Enables accent-aware ASR models.

✍️
பிழை–திருத்தம் இணைகள்
Error Correction Pairs

Crowd-sourced spelling and grammar errors paired with corrections. Vital for fine-tuning Tamil spell-check models.

👦
குழந்தை குரல் தரவு
Children's Speech

Age-annotated speech recordings of children (6–14) — essential for training models that understand non-adult pronunciation patterns.

நாமே கட்டுவோம்

Thamizharivu is not a product — it is a movement. Every teacher who records a lesson, every parent who annotates a story, every developer who commits code is building the future of Tamil AI.

🎙️
தரவு பங்களிப்பு — Contribute Data

Record 100 Tamil words on your phone. Upload via our simple web form. Your voice helps train the next generation of Tamil AI models.

  • Record Tamil phonemes & words
  • Annotate your regional accent
  • Donate public-domain text
💻
திறந்த மூலக் குறியீடு — Open Source

All models, datasets, and tools are released on GitHub under permissive licences. Fork, improve, and submit pull requests.

  • Fine-tune Tamil ASR models
  • Build Tamil NLP pipelines
  • Improve the spelling checker
🏫
பள்ளிகளுக்கு — For Schools

We partner with Tamil-medium schools and heritage language programs worldwide. Free access, Tamil-first curriculum integration, and teacher support.

  • Free school accounts
  • Classroom progress dashboards
  • Teacher resource packs
🌐
புலம்பெயர் சமூகம் — Diaspora

Growing up outside Tamil Nadu? Our diaspora program helps second-generation Tamil children connect with their heritage language at their own pace.

  • Weekend school resources
  • Diaspora accent recognition
  • Parent & child learning kits