Teuken-7B-Instruct
Germany's sovereign open LLM - 7B trained from scratch on all 24 EU official languages.
About
Teuken-7B is the flagship model of OpenGPT-X, the German government-funded (BMWK, ~14M EUR) project to build European AI sovereignty. Developed by a consortium led by Fraunhofer IAIS and IIS with TU Dresden, DFKI, and the Jülich Research Center, it was trained from scratch on the JUWELS supercomputer across all 24 official EU languages, with roughly half the pre-training data non-English.
The instruct variant (v0.4) has been available since November 26, 2024, with a commercial-use variant under Apache 2.0; Deutsche Telekom's T-Systems has productized it for European businesses and public authorities. A research-only v0.6 track also exists with separate licensing.
Honest limits: as a 7B from-scratch model it cannot match frontier-scale systems on general benchmarks - its value is sovereignty, GDPR-compliant deployment, and balanced multilingual performance. Licensing differs per variant, so check the exact model card before commercial use.
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 — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 88; Trending score 68/100 from researcher assessment.