EuroLLM-9B-Instruct (2512)
EuroLLM (utter-project) · Providers EU
Laatst beoordeeld:
Soevereiniteit en hosting
No endpoint for this model is reachable right now (de-fra Gepland onderhoud). The region and prices describe the deployment, not current availability.
Samenvatting
EuroLLM 9B (eurollm-9b-instruct-2512) is een 9.15B fine-tune voor Europese talen. Het omvat de 24 officiële EU-talen plus Catalaans, Galicisch, Turks, Arabisch, Chinees, Japans, Koreaans, Russisch en Hindi, onder Apache-2.0 met een contextvenster van 32k tokens en een post-training recept overgenomen van de 22B-release.
Specificaties
- Provider
- EuroLLM (utter-project) (PT)
- Contextvenster
- 33k
- Parameters
- 9.15B
- Modaliteiten
- text
- Licentie
- apache-2.0 · licence
- Open weights
- yes
- Releasedatum
- 26 jan 2026
- Taken
- Vertaling, Meertalige ondersteuning, Samenvatten, Retrieval (RAG)
Talen
Sterk: en, de, es, fr, it, pt, pl, nl, sv, cs, el, ro, fi, da, uk
Voldoende: hu, bg, sk, sl, lt, lv, et, hr, ga, mt, ca, tr, ar, zh, ja, ko, ru, hi
Sterk
- en
- de
- es
- fr
- it
- pt
- pl
- nl
- sv
- cs
- el
- ro
- fi
- da
- uk
Voldoende
- hu
- bg
- sk
- sl
- lt
- lv
- et
- hr
- ga
- mt
- ca
- tr
- ar
- zh
- ja
- ko
- ru
- hi
Niet geverifieerd: de provider heeft deze claim niet gedocumenteerd. — a language is listed as strong only where a source claims it; the card links the evidence where one exists.
Wanneer te gebruiken
Sterke punten
- explicit coverage of the 24 official EU languages
- Apache-2.0 licence with no user threshold
- runs on a single mid-range accelerator
- long-context extension phase applied in training
Beperkingen
- 32k-token context window is small next to current long-context models
- reasoning and coding quality below same-size commercial models
- the project publishes no independent third-party evaluation
Veelgestelde vragen
Which languages does EuroLLM-9B-Instruct (2512) cover?
The card records 15 strong languages (en, de, es, fr, it, pt, pl, nl, sv, cs, el, ro, fi, da, uk) and 18 more as supported (hu, bg, sk, sl, lt, lv, et, hr, ga, mt, ca, tr, ar, zh, ja, ko, ru, hi). Its summary says it was trained for the 24 EU official languages plus Catalan, Galician, Turkish, Arabic, Chinese, Japanese, Korean, Russian and Hindi — European language coverage rather than frontier reasoning. Source: https://huggingface.co/utter-project/EuroLLM-9B-Instruct-2512.
How many parameters does EuroLLM-9B-Instruct (2512) have, and what is its context window?
The card records 9.15B (safetensors total 9,152,319,488) (source: https://huggingface.co/api/models/utter-project/EuroLLM-9B-Instruct-2512) with a context window of 32768 tokens (source: https://huggingface.co/utter-project/EuroLLM-9B-Instruct-2512/blob/main/config.json).
What licence is EuroLLM-9B-Instruct (2512) under, and does it permit commercial use?
Apache License 2.0 (https://www.apache.org/licenses/LICENSE-2.0). The card lists "Apache-2.0 licence with no user threshold" among its strengths, so commercial serving is permitted under that licence — the licence text governs.
Is EuroLLM-9B-Instruct (2512) served on LLM EU?
It has no endpoint in the catalogue, so it is catalogue only — not hosted on LLM EU yet, with no region and no retention to report. A card whose status is "listed" is a catalogue entry, not a running deployment.
Has EuroLLM-9B-Instruct (2512) been evaluated, and by whom?
LLM EU has run no evaluation of it: the card carries no benchmark table, and its limits state that the project publishes no independent third-party evaluation and that reasoning and coding quality is below same-size commercial models. Source: https://huggingface.co/utter-project/EuroLLM-9B-Instruct-2512.