RUNLOCALAIv38
->Will it run?Best GPUCompareTroubleshootStartLearnPulseModelsHardwareToolsBench
Run check
RUNLOCALAI

Independently operated catalog for local-AI hardware and software. Hand-written verdicts. Source-cited claims. Reproducible commands when we have them.

OP·Fredoline Eruo
DIR
  • Models
  • Hardware
  • Tools
  • Benchmarks
TOOLS
  • Will it run?
  • Compare hardware
  • Cost vs cloud
  • Choose my GPU
  • Prompting kits
  • Quick answers
REF
  • All buyer guides
  • Learn local AI
  • Methodology
  • Glossary
  • Errors KB
  • Trust
EDITOR
  • About
  • Author
  • How we make money
  • Editorial policy
  • Contact
LEGAL
  • Privacy
  • Terms
  • Sitemap
MAIL · MONTHLY DIGEST
Get monthly local AI changes
Monthly recap. No spam.
DISCLOSURE

Some links on this site are affiliate links (Amazon Associates and other first-class retailers). When you buy through them, we earn a small commission at no extra cost to you. Affiliate links do not influence our verdicts — there are cards we rate highly that we don't have affiliate relationships with, and cards that sell well that we refuse to recommend. Read more →

© 2026 runlocalai.coIndependently operated
RUNLOCALAI · v38
BLK · TURKISH MODELSTürkçe açık ağırlıklı modeller

Turkish open-weight LLMs

Every Turkish-tuned open-weight model we’ve verified on Hugging Face. Operator-curated catalog — specs from published model cards, no vendor marketing copy. Updated as new releases ship from Trendyol, YTU CE COSMOS, VNGRS, and the broader Turkish AI community.

Models curated
20
Organizations
11
Commercial OK
15/20
Benchmarked
3/20

Why a separate Turkish catalog?

English-trained LLMs handle Turkish poorly because of how BPE tokenization interacts with Turkish’s agglutinative morphology. A single Turkish word (e.g., evlerimizdekilerden) can fragment into 8+ tokens on a Llama 3 tokenizer, vs ~3 on Qwen 3 (which trained for 119 languages) and ~1 on a purpose-built Turkish tokenizer. More tokens means more compute, more VRAM, and degraded attention quality on the same logical sentence.

That’s why a Turkish-from-scratch model (Kanarya), or a continued-pretrained Turkish fine-tune (the YTU CE COSMOS family, Trendyol’s flagship), often outperforms a larger English-base model on Turkish-language tasks despite having a fraction of the parameters. The Turkish-language community has its own ecosystem, and the catalog below covers the active releases.

Each entry links to the model page on /models with the same depth of editorial we give every model. Where we’ve run benchmarks (TurkishMMLU specifically), the score appears next to the model. Where we haven’t yet, it says “not yet tested” — we don’t fake numbers.

FAM · OTHER

Other / from-scratch

6 models
Omni 31B Turkish Reasoning
31B params · community
not yet tested

31B-parameter Turkish-tuned reasoning model with i1-imatrix quantizations by mradermacher. Designed for step-by-step problem solving in Turkish. Highest download count among large Turkish-tuned models.

License
unknown
Context
32K
Mihenk LLM v2 35B (Turkish Financial)
35B params · emircansevdi
not yet tested

35B MoE (3B active) tuned specifically for Turkish financial-services text — bank statements, investment research, accounting terminology. Niche-cluster model from the Turkish fintech community.

License
unknown
Context
32K
Kanarya 2B
2B params · asafaya
not yet tested

Turkish-from-scratch language model trained by Ali Safaya (Koç University researcher). Named after the kanarya (Turkish for 'canary'). Trained on 250+ GB of Turkish text including Wikipedia, news, and books.

License
Apache-2.0 · COMMERCIAL OK
Context
2K
VBART Large (Turkish Summarization)
0.4B params · VNGRS
not yet tested

Turkish BART-style sequence-to-sequence model fine-tuned specifically for summarization. Not a chat model — purpose-built for input-document → Turkish-summary pipelines.

License
Apache-2.0 · COMMERCIAL OK
Context
1K
Kanarya 750M
0.75B params · asafaya
not yet tested

Smaller Kanarya variant — 750M parameters. Runs on CPU or 4GB GPU comfortably. Useful for low-resource Turkish text classification, embeddings, or completion tasks where latency matters more than quality.

License
Apache-2.0 · COMMERCIAL OK
Context
2K
Turkish GPT-2 Large
0.7B params · ytu-ce-cosmos
not yet tested

GPT-2 Large architecture trained from scratch on Turkish. Reference baseline for measuring how much modern instruction-tuned models actually improve on the GPT-2 era.

License
MIT · COMMERCIAL OK
Context
1K
FAM · MISTRAL

Mistral-based

5 models
Kumru 2B
2.4B params · VNGRS
not yet tested

Kumru 2B is a compact Turkish text-generation model from VNGRS. The Hugging Face config reports a Mistral-family architecture with an 8K context window, and the public Ollama build makes it a practical edge-speed Turkish

License
Apache-2.0 · COMMERCIAL OK
Context
8K
Turkcell LLM 7B v1
7.4B params · Turkcell
not yet tested

Turkcell LLM 7B v1 is an Apache-2.0 Turkish text-generation model built on a Mistral architecture. The measured Ollama artifact uses a RefinedNeuro GGUF distribution of the public Turkcell model.

License
Apache-2.0 · COMMERCIAL OK
Context
32K
Turkish Mistral 7B Instruct v0.2
7B params · ytu-ce-cosmos
not yet tested

Mistral 7B v0.2 continued-pretrained on Turkish data + instruction-tuned. The 32K context window makes it the best Turkish open-weight model for long-document work today.

License
Apache-2.0 · COMMERCIAL OK
Context
32K
Mistral Turkish v2 (brooqs)
7.2B params · brooqs
not yet tested

Mistral Turkish v2 is a public Ollama-distributed Turkish Mistral variant. The upstream Hugging Face repository was not publicly accessible during intake, so commercial-use status is kept conservative until the source li

License
unknown
Context
8K
Malhajar Mistral 7B Turkish
7.2B params · malhajar
not yet tested

Malhajar Mistral 7B Turkish is an Apache-2.0 Mistral 7B Instruct v0.2 Turkish fine-tune. The benchmarked Ollama tag is a koezgen quantized distribution of the public malhajar model.

License
Apache-2.0 · COMMERCIAL OK
Context
32K
FAM · GEMMA

Gemma-based

4 models
Turkish Gemma 9B T1
9B params · ytu-ce-cosmos
not yet tested

YTU's Turkish-tuned Gemma 2 9B model. The highest community-rated Turkish-language LLM on Hugging Face by likes-to-downloads ratio as of May 2026. Continued pre-training on Turkish corpora followed by instruction tuning.

License
Gemma · COMMERCIAL OK
Context
8K
Trendyol LLM Asure 12B
11.8B params · Trendyol
58.9/100 (Q4_K_M)

Trendyol LLM Asure 12B is a Gemma 3 based multimodal instruct model for Turkish and English business workflows. The public Ollama build used in local testing is the alibayram GGUF distribution.

License
Gemma · COMMERCIAL OK
Context
128K
Gemma 4 Turkish 26B (4B active)
26B params · esokullu
not yet tested

Gemma 4 26B MoE (4B active params) pruned and Turkish-tuned. The largest Turkish-tuned open-weight model on HF as of May 2026. MoE architecture means it loads 26B of weights but runs at 4B-active speed.

License
Gemma · COMMERCIAL OK
Context
128K
YTU Turkish Gemma 9B v0.1
9.2B params · ytu-ce-cosmos
not yet tested

YTU Turkish Gemma 9B v0.1 is a Gemma 2 based Turkish instruction model from the YTU CE COSMOS ecosystem. The benchmarked Ollama tag is an alibayram GGUF distribution of that public model.

License
Gemma · COMMERCIAL OK
Context
8K
FAM · LLAMA

Llama-based

4 models
Turkish Llama 8B Instruct v0.1
8B params · ytu-ce-cosmos
11.0/100 (Q4_K_M)

Llama 3 8B continued pre-trained on Turkish corpora, then instruction-tuned for Turkish chat. YTU CE COSMOS group's most-downloaded Llama variant. GGUF builds available — drops into Ollama directly.

License
Llama-3-Community · COMMERCIAL OK
Context
8K
Trendyol LLM 7B Chat v0.1
7B params · Trendyol
not yet tested

Turkish-tuned chat model released by Trendyol, Turkey's largest e-commerce platform. Built on Llama 2 7B, fine-tuned on Turkish customer-service style dialogues plus general Turkish chat. The first major Turkish LLM from

License
Trendyol-LLM-Community
Context
4K
Cosmos Llama 3 8B Turkish
8B params · ytu-ce-cosmos
not yet tested

YTU CE COSMOS's Llama 3 8B Turkish instruction-tuned variant. Follow-up to the original Turkish-Llama-8b that uses the Llama 3 base instead of Llama 2 — better tokenizer, stronger zero-shot Turkish.

License
Llama-3-Community · COMMERCIAL OK
Context
8K
Trendyol LLM 7B Base v0.1
7B params · Trendyol
not yet tested

Base (non-chat) variant of Trendyol's 7B Turkish LLM. The chat sibling is the more popular pick; this base version is for operators building their own instruction-tuning pipeline on top of Trendyol's pre-training.

License
Trendyol-LLM-Community
Context
4K
FAM · QWEN

Qwen-based

1 model
Qwen 3.5 2B Turkish SFT
2B params · Tuguberk
not yet tested

Qwen 3.5 2B base with supervised fine-tuning on Turkish instruction-following data. Recent community fine-tune (early 2026) that bridges Qwen 3.5's strong multilingual base with Turkish-specific chat capability.

License
Apache-2.0 · COMMERCIAL OK
Context
32K
BENCHMARKS · TURKISH-MMLU

We’re running TurkishMMLU on every model above

TurkishMMLU is a 900-question, 9-subject Turkish-language benchmark for general knowledge and reasoning. Scores appear next to each model as our compute finishes. Each result links to the raw test-run Gist for verification — same trust apparatus as our /benchmarks/coding (HumanEval+) leaderboard.

DON’T SEE YOUR MODEL?

Turkish AI community moves fast. If we missed a release, tell us.

Every catalog entry above passes a discovery gate (downloads ≥ 50 on Hugging Face, updated within the last 18 months, actually Turkish — not multilingual with a Turkish token in the slug). If a model meets those criteria and isn’t listed, point us to the HF repo and we’ll add it on the next sweep.