Open Slovenian language models
Curated list of open models for Slovenian: LLMs, translation, speech, encoders, NER, benchmarks. Suggestions welcome.
Text Generation • 12B • Updated • 3.14k • 8Note LLM · CJVT / UL FRI · reference open Slovenian model (GaMS family).
cjvt/GaMS3-12B-Instruct-Lex
Text Generation • 13B • Updated • 1.07kNote LLM · CJVT / UL FRI · lexically improved GaMS3 instruct.
cjvt/GaMS3-12B
Text Generation • 12B • Updated • 438Note Base LLM · CJVT / UL FRI.
cjvt/GaMS-27B-Instruct
Text Generation • 27B • Updated • 301 • • 1Note LLM · CJVT / UL FRI.
cjvt/GaMS2-27B-Instruct
852k • Updated • 82Note LLM · CJVT / UL FRI.
cjvt/GaMS-9B-Instruct
Text Generation • 9B • Updated • 1.95k • • 2Note LLM · CJVT / UL FRI.
cjvt/GaMS-9B-Instruct-Nemotron
Text Generation • 9B • Updated • 65 • • 1Note LLM · CJVT / UL FRI.
cjvt/GaMS-27B-Instruct-Nemotron
Text Generation • 27B • Updated • 50 • • 5Note LLM · CJVT / UL FRI.
cjvt/GaMS-2B-Instruct
Text Generation • 3B • Updated • 401 • 2Note Small LLM · CJVT / UL FRI.
cjvt/GaMS-1B-Chat
Text Generation • 2B • Updated • 50 • 3Note Small LLM · CJVT / UL FRI (2024).
GaMS-Beta/SVILA-1-12B
Text Generation • 12B • Updated • 58 • 1Note LLM · GaMS-Beta.
GaMS-Beta/SVILA-1-4B
Text Generation • 4B • Updated • 26Note Small LLM · GaMS-Beta.
cjvt/GaMS3-12B-Multimodal
Text Generation • 12B • Updated • 14Note Vision-language model · GaMS-Beta.
texdata/sraka-27b-experimental
Text Generation • 27B • Updated • 6Note LLM, 27B · MediaAtlas · Slovenian chat, reasoning, tool calling. Gated (request access), non-commercial.
texdata/Sokol-4B-SLO
Text Generation • 5B • Updated • 779Note Small on-device LLM (4B) · MediaAtlas · chat, EN↔SL, tool calls.
texdata/Qwen3.6-35B-A3B-Slovenian-GGUF
Text Generation • 36B • Updated • 447 • 3Note LLM, 35B MoE · MediaAtlas · Slovenian CPT+SFT, GGUF.
texdata/Qwen3.6-35B-A3B-Tool-Slo-GGUF
Text Generation • 36B • Updated • 361Note LLM, 35B MoE · MediaAtlas · tool / function calling.
texdata/Vran-27B-SLO-BioMed-Research
Text Generation • 27B • Updated • 2Note Biomedical research LLM (27B) · MediaAtlas · research only.
zID4si/Zlatorog-12B-Instruct-Beta
Text Generation • 12B • Updated • 2 • 2Note LLM · Zlatorog (beta).
zID4si/Zlatorog-30B-A3B-Instruct-Experimental
Text Generation • 31B • UpdatedNote LLM, 30B MoE · Zlatorog (experimental).
zlatorog/Zlatorog-30B-MoE-v1-beta
31B • Updated • 1Note LLM, 30B MoE · Zlatorog (beta, gated).
SemantikaEU/Micka-gams-2b
2B • Updated • 7Note Small LLM · Semantika.
sambanovasystems/SambaLingo-Slovenian-Chat
Text Generation • 7B • Updated • 56 • 30Note LLM · SambaNova (2024).
sambanovasystems/SambaLingo-Slovenian-Base
Text Generation • Updated • 165 • 27Note Base LLM · SambaNova (2024).
gordicaleksa/SlovenianGPT
7B • Updated • 8 • 11Note LLM (2024).
xdatasi/antares-7b-slovenian
Text Generation • 7B • Updated • 4Note LLM, 7B (2024).
RomanOrac/llama-2-7b-slovenian
Text Generation • Updated • 18 • 1Note LLM, 7B (2023).
alphaedge-ai/Qwen3.5-4B-slv-32768
Text Generation • 4B • Updated • 19Note Qwen3.5-4B with vocabulary trimmed to Slovenian (smaller, same weights otherwise) · alphaedge-ai publishes ~200 such variants.
klemenKM/pravnik-v01
Text Generation • Updated • 6Note Legal LLM for Slovenian law (proof of concept, QLoRA).
labpt/ContRAG-GaMS3
Text Generation • Updated • 7Note Legal NLI fine-tune of GaMS3-12B.
klemenp950/GaMS-9B-medical-email-Slovene
Text Generation • Updated • 11Note Medical e-mail fine-tune of GaMS-9B (LoRA).
klemenp950/gams-triage-summary-v8
2B • Updated • 7Note Triage-summary fine-tune of GaMS (student project).
alenkaz/GaMS3-12B-v6_subtitles_no-identicals-datasets
Text Generation • Updated • 13Note Subtitles fine-tune of GaMS3-12B (student project).
lea-vodopivec7/gams-2b-finetuned-kks-V1
Text Generation • Updated • 11Note Fine-tune of GaMS-2B-Instruct (student project).
cjvt/GaMS3-12B-GRPO-Translator
12B • Updated • 69Note EN⇄SL translation LLM · CJVT / UL FRI.
GaMS-Beta/GaMS-9B-SFT-Translator
Text Generation • 9B • Updated • 14 • 1Note EN⇄SL translation LLM · GaMS-Beta.
texdata/Qwen3.6-27B-slo-med-mt-GGUF-Q8
Translation • 27B • Updated • 236Note Medical EN⇄SL translation (27B) · MediaAtlas.
texdata/Qwen3.6-27B-slo-med-mt
Translation • 27B • Updated • 31Note Medical EN⇄SL translation (27B, merged, vision) · MediaAtlas.
texdata/Qwen3.6-35B-A3B-slo-med-mt
Translation • Updated • 35Note Medical EN⇄SL translation (35B MoE, LoRA) · MediaAtlas.
allegro/BiDi-eng-slv
Translation • 0.2B • Updated • 5Note EN⇄SL translation · Allegro.
texdata/nemotron-3.5-asr-streaming-slovenian
Automatic Speech Recognition • 0.6B • Updated • 237 • 3Note Speech recognition, streaming · MediaAtlas.
anton-l/wav2vec2-large-xlsr-53-slovenian
Automatic Speech Recognition • Updated • 272kNote Speech recognition (wav2vec2 XLSR), most downloaded SL ASR.
yuriyvnv/parakeet-tdt-0.6b-slovenian
Automatic Speech Recognition • Updated • 45 • 2Note Speech recognition (Parakeet).
yuriyvnv/whisper-large-v3-slovenian
Automatic Speech Recognition • 2B • Updated • 48 • 1Note Speech recognition (Whisper large-v3).
samolego/whisper-small-slovenian
Automatic Speech Recognition • 0.2B • Updated • 89 • 3Note Speech recognition (Whisper small).
aklemen/conformer-sloveniangpt-baseline
8B • Updated • 7Note Speech recognition (Conformer).
BuzzASR/slovenian
Automatic Speech Recognition • 2B • Updated • 129Note Speech recognition.
facebook/wav2vec2-base-10k-voxpopuli-ft-sl
Automatic Speech Recognition • Updated • 12Note Speech recognition (VoxPopuli) · Meta.
blko/whisper-base-sl-artur-full-ft
Automatic Speech Recognition • 72.6M • Updated • 18Note Speech recognition (Whisper base) trained on the ARTUR corpus.
blko/whisper-tiny-sl-artur-full-ft
Automatic Speech Recognition • 37.8M • Updated • 6Note Speech recognition (Whisper tiny) trained on the ARTUR corpus.
aklemen/conformer-gams-9b-lora-r16-a32
9B • Updated • 10Note Speech-LLM research: Conformer encoder + GaMS-9B (LoRA).
texdata/Gepard-Slovenian-TTS
Text-to-Speech • 0.6B • Updated • 89 • 3Note Text-to-speech · MediaAtlas.
texdata/Maya1-Slovenian-TTS
Text-to-Speech • Updated • 121 • 1Note Text-to-speech · MediaAtlas.
speaches-ai/piper-sl_SI-artur-medium
Text-to-Speech • Updated • 20Note Text-to-speech (Piper, Artur voice).
EMBEDDIA/sloberta
Fill-Mask • Updated • 4.89k • • 5Note Encoder · SloBERTa.
EMBEDDIA/crosloengual-bert
Fill-Mask • Updated • 4.11k • • 7Note Encoder · Croatian/Slovenian/English BERT.
classla/xlm-r-slobertic
Fill-Mask • Updated • 575Note Encoder · CLASSLA.
LHRS-UM-FERI/SIGMA-BERT-base
Fill-Mask • 0.1B • Updated • 153Note Encoder (ModernBERT) · UM FERI.
LHRS-UM-FERI/SIGMA-BERT-large
Fill-Mask • 0.4B • Updated • 77Note Encoder (ModernBERT) · UM FERI.
HPLT/hplt_bert_base_2_0_slv-Latn
Fill-Mask • 0.2B • Updated • 27Note Encoder · HPLT.
joelniklaus/legal-slovenian-roberta-base
Fill-Mask • Updated • 192Note Legal-domain encoder.
alphaedge-ai/bge-m3-slv-32768
Sentence Similarity • 0.3B • Updated • 67Note BGE-M3 embeddings with vocabulary trimmed to Slovenian · alphaedge-ai.
cjvt/t5-sl-large
Updated • 137 • 3Note Slovenian T5 · CJVT.
stanfordnlp/stanza-sl
Token Classification • Updated • 1.05kNote Tagging, lemmas, parsing · Stanza.
spacy/sl_core_news_trf
Token Classification • Updated • 9Note spaCy pipeline (transformer).
clarin-pl/combo-nlp-xlm-roberta-base-slovenian-ssj-ud2.17
Token Classification • Updated • 11Note UD parsing (COMBO, SSJ).
LHRS-UM-FERI/SIGMA-BERT-base-ner
Token Classification • 0.1B • Updated • 32Note Named-entity recognition · UM FERI.
martinkorelic/rudar-mmbert-slv-ner
Token Classification • 0.3B • Updated • 272Note Named-entity recognition · Rudar.
HUMADEX/slovenian_medical_ner
Token Classification • 0.1B • Updated • 17Note Medical named-entity recognition.
cjvt/sloberta-si-nli
Text Classification • Updated • 13Note Natural language inference · CJVT.
cjvt/sloberta-trendi-topics
Text Classification • Updated • 149Note News topic classification · CJVT.
maticzav/slovenian-news-topic-classifier
Text Classification • 0.3B • Updated • 9Note News topic classification.
IMSyPP/hate_speech_slo
Text Classification • Updated • 19Note Hate-speech classification · IMSyPP.
cjvt/SloBERTa-slo-word-spelling-annotator
Fill-Mask • Updated • 20 • 1Note Spelling annotation · CJVT.
texdata/t5-sl-medical-gec
Text Generation • 76.8M • Updated • 205Note Medical grammatical error correction · MediaAtlas.
sharpsy/slovene-morphological-segmentation-sloberta
0.1B • Updated • 11Note Morphological segmentation.
janrobas/Qwen2.5-0.5B-Slovenian-PersonDataExtractor
0.5B • Updated • 10Note Person-data extraction (0.5B).
cjvt/slovenian-llm-eval
Viewer • Updated • 271k • 112Note Benchmark · Slovenian LLM evaluation suite · CJVT.
texdata/med-slo-sft
Preview • Updated • 19Note Dataset · Slovenian medical SFT corpus · MediaAtlas.