How do you build Tamil AI apps with Plugsky?

Key facts

Tamil script	Left-to-right abugida with attached vowel signs and pulli marks
Diglossia	Written and spoken Tamil differ enough to need separate evaluations
Tokenisation	Agglutinated forms fragment into multiple tokens — measure with the Plugsky token calculator
API compatibility	OpenAI-compatible POST https://api.plugsky.com/v1/chat/completions — chat, streaming, JSON mode and function calling
Models	30+ models behind one endpoint, from free tiers to frontier reasoning
Free tier	Free plan with plugsky-micro and plugsky-lite, no card required
Deployment	Plugsky cloud, your VPC, on-prem and air-gapped options
Product status	Chat, streaming, JSON mode, function calling, embeddings, RAG and agents live; audio, images, batch and fine-tuning coming soon

TL;DR

• Tamil runs on the standard chat endpoint — one base_url change.
• NFC-normalise the script so vowel signs stay attached.
• Written and spoken Tamil need separate prompt and eval sets.
• Romanised Tamilish needs a transliteration path for search.
• plugsky-embed-multilingual covers Tamil and English retrieval.

How it works, step by step

1. Create a Plugsky API key on the free plan (no card) and set base_url to https://api.plugsky.com/v1.
2. Send a small set of real Tamil prompts to /v1/chat/completions and compare output across two or three models.
3. Count tokens for those prompts with the Plugsky token calculator and set chunk sizes that fit your model context.
4. Normalise text before indexing or prompting: NFC-normalise the Tamil script and standardise loanword spelling.
5. For RAG, embed with plugsky-embed-multilingual and test cross-language queries alongside Tamil-only queries.
6. Score candidate models on a Tamil gold set with native-speaker review, then cut production traffic over.

Try it yourself

Open the RAG sandbox →

How Plugsky handles Tamil text

Tamil is written left to right in the Tamil script, an abugida where vowel signs and pulli marks attach to consonants.

Written Tamil and spoken Tamil are markedly different (diglossia), and Sri Lankan and Singapore Tamil add vocabulary and spelling variation.

Evaluate with a Tamil speaker on written and spoken samples; check script accuracy and register consistency, since the gap between the two forms is large.

Tokenisation and cost in Tamil

Tamil agglutinates heavily and the script uses fewer consonant letters than related Indic scripts, so tokenisers often split words into several pieces; transliterated Tamilish tokenises very differently.

• NFC-normalise the script so combining marks stay attached.
• Test written and spoken Tamil as separate sets.
• Handle English and Sanskrit loanwords in one glossary.
• Count tokens on real messages, not dictionary words.

Tamil retrieval and RAG

Index written Tamil documents and test spoken-style and transliterated queries, because users type closer to speech than to documents.

• Use plugsky-embed-multilingual for Tamil and English corpora.
• NFC-normalise the script before embedding.
• Add transliteration for Romanised queries.
• Evaluate Sri Lankan and Singapore variants separately.

Code example: a Tamil request

Point your existing OpenAI client at https://api.plugsky.com/v1 and pass Tamil text in the content field — no language flag and no separate endpoint. Streaming, JSON mode and function calling keep the same request shapes.

client = OpenAI(base_url="https://api.plugsky.com/v1", api_key=os.environ["PLUGSKY_API_KEY"])

client.chat.completions.create(model="plugsky-pro", messages=[{"role": "user", "content": "இந்த ஒப்பந்தத்தை தமிழில் மூன்று புள்ளிகளில் சுருக்கவும்."}])

Start on the free plan with plugsky-micro and plugsky-lite, then compare paid models on a Tamil gold set before cutover. See the docs for request details.

For production, log the model name and your normalisation settings with each request, and re-run the Tamil gold set whenever either changes — language quality regressions usually come from prompt or preprocessing drift, not from the model alone.

Honest comparison

Capability	Plugsky	Tamil workflow today	Building in-house
API compatibility	OpenAI-compatible — change base_url and model name	Varies by provider and SDK	Full rewrite
Tamil text handling	NFC script handling with written and spoken separation	Depends on provider tokeniser and prompt hygiene	You build normalisation, segmentation and evals
Token budget	Fixed tokeniser per model; measure with the Plugsky token calculator and chunk to fit	Varies by provider and model	You host and tune each tokeniser
Multilingual retrieval	plugsky-embed-multilingual available for cross-language RAG	Often needs a separate embedding vendor	You serve and maintain embeddings
Sovereignty	Cloud, VPC, on-prem and air-gapped with residency options	Usually US/EU public endpoints	You own the full stack

Frequently asked questions

Can Plugsky handle Tamil text?

Yes. The API accepts UTF-8 Tamil input on the OpenAI-compatible chat endpoint; output quality depends on the model, so compare two or three on your own prompts before choosing.

How do I estimate token usage for Tamil?

Tamil agglutinates heavily and the script uses fewer consonant letters than related Indic scripts, so tokenisers often split words into several pieces; transliterated Tamilish tokenises very differently. Use the token calculator at /tools/llm-token-calculator before sizing context windows or chunk lengths.

Written or spoken Tamil?

Documents and formal content use written Tamil; chat is closer to spoken forms. Evaluate both because the gap is large.

Can users type Tamilish in Roman script?

Yes, but it needs its own handling. Add transliteration for search and test Roman queries against Tamil documents with plugsky-embed-multilingual.

Is there a multilingual embedding model?

Yes — plugsky-embed-multilingual is part of the 30+ model catalogue and is built for cross-language retrieval. Keep one embedding model per vector collection.

Can I keep data in my region?

Plugsky supports cloud, VPC, on-prem and air-gapped deployment with data-residency options; confirm your requirements with the docs and the enterprise team.

How do I migrate an existing app?

Change base_url to https://api.plugsky.com/v1 and map the model name. Streaming, JSON mode, function calling and embeddings keep the same request shapes.

Is there a free plan?

Yes — the free plan includes two free models, plugsky-micro and plugsky-lite, with no card. A 14-day full-access trial unlocks the paid catalogue.