Is it good enough?
This page is a projection of the one entry record, the Reliability factor that Use covers. The full verdict is set by all four factors together, floor-weighted so the weakest caps the whole.
- AssessUse & modify + Transparency
- ImplementData control + Reliability
- UseReliability
- SupportTransparency
Capabilities & modalities
EuroLLM is a text-only multilingual language model for European-language generation, assistant
tasks and machine translation. Per aggregated benchmarks and the technical report
(ev-tech-report, ev-moonlight) it leads its size on multilingual EU-language tasks: on par
with Gemma-2-9B on multilingual benchmarks, ahead on WMT24++ translation, and matching Mistral-7B
on English. At 9B it remains below frontier on raw single-language reasoning. Choose EuroLLM for
broad, even EU-language coverage from a European, Apache-2.0 model when class-topping English
reasoning is not the priority.
Context window & long-context behaviour
Both EuroLLM-9B and EuroLLM-1.7B use a 4,096-token context window. This is short by current standards, so design around it:
- Chunk long inputs and use retrieval instead of stuffing large documents into the prompt.
- Budget the 4k across system prompt, few-shot examples, input and output.
Treat 4k as the architectural ceiling.
Prompt format & chat template
The Instruct variants ship a chat template in their tokenizer configuration (BOS token id 1, EOS token id 4). Use it instead of hand-rolling role markers:
messages = [
{"role": "system", "content": "You are a helpful multilingual assistant."},
{"role": "user", "content": "Traduis en portugais : ..."},
]
inputs = tok.apply_chat_template(messages, tokenize=True,
add_generation_prompt=True, return_tensors="pt")
Base checkpoints have no chat template. They are plain text-completion models and should not be prompted as chat assistants.
Language coverage
Language coverage is EuroLLM's defining strength. The models are trained from scratch for multilingual EU coverage: all 24 official EU languages - Bulgarian, Croatian, Czech, Danish, Dutch, English, Estonian, Finnish, French, German, Greek, Hungarian, Irish, Italian, Latvian, Lithuanian, Maltese, Polish, Portuguese, Romanian, Slovak, Slovenian, Spanish, Swedish - plus roughly 11 additional strategic languages: Arabic, Catalan, Chinese, Galician, Hindi, Japanese, Korean, Norwegian, Russian, Turkish and Ukrainian.
EU-language coverage is EuroLLM's design target, where English-centric open models treat non-English behaviour as incidental, and translation quality (WMT24++) is a documented strength. The caveat is that with 35 languages sharing a 9B (or 1.7B) budget, quality is uneven: the smaller and lower-resource languages trail the majors. Evaluate each target language for your task instead of assuming uniform quality.
How this scores
The ownership factor this domain covers, drawn from the one entry record.
ReliabilityIs it reliable and good enough for the job?
ModerateRuns dependably, but per-language safety coverage is thin and uneven, so you supply the safety layer.
Sources
The same evidence records as the entry sheet. Read means the text was verified; unverified means it is known to exist but not yet read.