How do you build Turkish AI apps with Plugsky?

Key facts

Turkish grammar	Agglutinative: suffix chains stack onto stems, producing long words
Casing	Dotted and dotless i (İ, ı) break naive lowercase and uppercase operations
Tokenisation	Suffix chains split into many subwords — measure with the token calculator
Register	Formal siz and informal sen change verb forms — keep one per surface
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
Product status	Chat, streaming, JSON mode, function calling, embeddings, RAG and agents live; audio, images, moderation, files, batch, fine-tuning, assistants and responses coming soon

TL;DR

• Turkish runs on the standard endpoint with one base_url change.
• Agglutination raises tokens per word — measure real Turkish text.
• Use Turkish-aware casing for the dotted and dotless i.
• Pin siz or sen per surface and state it in the system prompt.
• plugsky-embed-multilingual covers Turkish and English retrieval.

How it works, step by step

1. Create a free Plugsky key and set base_url to https://api.plugsky.com/v1.
2. Send real Turkish text to /v1/chat/completions and compare two or three models.
3. Measure tokens with the token calculator and add headroom for suffix-heavy text.
4. Check your search and prompt code for Turkish casing bugs around İ, ı, I and i.
5. Choose siz or sen per surface and write it into the system prompt.
6. For RAG, embed with plugsky-embed-multilingual and run a Turkish gold set with native review.

Try it yourself

Open the prompt optimizer →

How Plugsky handles Turkish text

Turkish needs no special endpoint: UTF-8 text goes in, text comes back. The language-specific work is split between casing and morphology. Turkish has both a dotted and a dotless i — İ/i and I/ı — and the default Unicode lowercase rule maps I to i, not to ı. Code that lowercases text without Turkish rules silently corrupts words in search, filtering and matching.

Formality is the second decision. siz and sen change verb endings across a sentence, and mixing them inside a flow is the most visible quality failure for Turkish readers.

Tokenisation and cost in Turkish

Turkish can stack several suffixes onto one stem, so words that would be phrases in English become single long tokens — and tokenisers split them into multiple subwords. Case, tense, person and negation can all appear in one word, and vowel harmony governs how the endings look.

• Measure on real product text, not on dictionary words.
• Leave more context headroom than you would for English.
• Keep suffixes attached for display; stem only for search keys.
• Re-measure whenever you change models.

Turkish retrieval and RAG

One multilingual collection built with plugsky-embed-multilingual serves Turkish documents with Turkish or English queries. Keyword search needs care: agglutinated forms produce many surface variants, so stemming or lemmatisation helps alongside embeddings, and casing must follow Turkish rules.

• Apply Turkish-aware casing before indexing or matching.
• Add stemming for keyword search on top of embeddings.
• Test English queries against Turkish documents explicitly.
• Keep one embedding model per collection.

Code example: a Turkish request

Point your OpenAI client at https://api.plugsky.com/v1; only the content and system prompt change.

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": "system", "content": "Türkçe yanıt ver, nazik bir dil kullan ve siz hitabını koru"}, {"role": "user", "content": "Bu sözleşmeyi üç maddede özetle"}])

Start on the free plan with plugsky-micro and plugsky-lite, then compare paid models on a Turkish gold set. See the docs for the API reference.

Honest comparison

Capability	Plugsky	Turkish apps today	Building in-house
API compatibility	OpenAI-compatible — change base_url and model name	Varies by provider and SDK	Full rewrite
Turkish text handling	UTF-8 with suffix-aware chunking and siz/sen prompt control	Depends on provider tokeniser and prompt hygiene	You build normalisation and evals
Token budget	Fixed tokeniser per model; expect more tokens per word and chunk accordingly	Varies by provider and model	You host and tune each tokeniser
Multilingual retrieval	plugsky-embed-multilingual for Turkish and English 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 Turkish text?

Yes. The API accepts UTF-8 Turkish input on the OpenAI-compatible chat endpoint. Compare two or three models on your own prompts, since quality varies by task.

Why do Turkish token counts run high?

Agglutination stacks suffixes onto stems, and tokenisers split those long words into several subwords. Use the token calculator on real text and leave context headroom.

What is the dotted and dotless i issue?

Turkish has İ/i and I/ı as distinct letters. Default Unicode lowercasing maps I to i, which is wrong for Turkish; use locale-aware casing in search, filters and prompts.

Should prompts use siz or sen?

siz for formal and B2B contexts, sen for consumer and community products. State the choice in the system prompt and keep it consistent.

Is there a multilingual embedding model?

Yes — plugsky-embed-multilingual is part of the 30+ model catalogue and supports Turkish and English retrieval from one collection.

Can I deploy in my own environment?

Yes. Plugsky supports cloud, VPC, on-prem and air-gapped deployment with residency options for regulated teams.

Is there a free plan?

Yes — plugsky-micro and plugsky-lite are free with no card, and a 14-day full-access trial covers the paid catalogue.