Occiglot
Europe's open LLM collective - 7B multilingual models for European language sovereignty.
About
Occiglot is a European academic research collective building open language models for Europe's languages and values. Its v0.1 release published ten intermediary 7B checkpoints - bilingual variants for English, German, French, Spanish, and Italian, plus a multilingual EU5 model - created through bilingual continual pre-training and instruction tuning starting from Mistral-7B.
The models were trained on roughly 700B additional multilingual tokens (with ~1B for instruction tuning), using compute from Hessian.AI's fortytwo supercomputer and the DFKI cluster. Everything is released under Apache 2.0 on Hugging Face, with a roadmap toward covering all 24 official EU languages.
Honest limits: these are early research checkpoints, not safety-aligned production models - the project itself notes they can generate problematic outputs. The release date here is era-level (early 2024), flagged honestly.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 6/10
Scored 6/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 88; Trending score 60/100 from researcher assessment.