Teuken Models by OpenGPT-X
Last refreshed 2026-10-05. Next refresh: weekly.
Details
About
Teuken is the multilingual large language model family from OpenGPT-X, a German research project. The models were developed by Fraunhofer, Forschungszentrum Jülich, TU Dresden, and DFKI, with funding from the German Federal Ministry of Economics and Climate Protection (BMWK). Teuken-7B is a 7B decoder-only model pretrained on 4 trillion tokens covering all 24 official European Union languages, aiming to give more stable results across those languages than English-centric models. Instruction-tuned versions are released for research and for commercial use, and they are meant to be used with the provided prompt template. The team says the models are not intended for math and coding tasks.
Decision facts
- Best fit
- codingmath-heavy prompts
- Capability starting point
- Teuken 7B Instruct with 4K context
- Lowest tracked input
- Not tracked
- Closest related family
- 360Zhinao
Current Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
Use when the workload needs 4K context and 7B parameters.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Teuken 7B Instruct | Use when the workload needs 4K context and 7B parameters. | 2024-10 | 4K context7B parameters | Current |
Release Timeline
1 release groupSpecifications(1 models)
| Model | Released | Context | Parameters |
|---|---|---|---|
| Teuken 7B Instruct | 2024-10 | 4K | 7B |
Available From(1 provider)
Popular comparisons in this family
- Codex 1 vs Teuken 7B Instruct11
- Mistral 7B v0.1 vs Teuken 7B Instruct8
- Llama 3.1 NemoGuard 8B Topic Control vs Teuken 7B Instruct7
- Mistral Medium 3 vs Teuken 7B Instruct7
- Llama 2 7B vs Teuken 7B Instruct6
- MiMo-V2.5 vs Teuken 7B Instruct6
- GPT-4 vs Teuken 7B Instruct6
- GPT-2 Medium vs Teuken 7B Instruct6
- Qwen3.5-9B vs Teuken 7B Instruct5
- Falcon 180B vs Teuken 7B Instruct5