Step 2 of 4 · ProjectsModels
SteuerLLM
28-billion-parameter language model for German tax law analysis, from the SteuerLLM paper.
CC BY-NC 4.0
What it does
A 28-billion-parameter language model adapted to German tax law, built by expanding a Mistral Small base and training on tax-filtered web data and synthetic question-answer pairs from German statutes such as the EStG, AO and KStG. Served locally with Transformers, vLLM or SGLang; the model card links the paper, a GitHub repository and a demo. CC BY-NC 4.0, research and academic use only.
Licence: CC BY-NC 4.0. Check its terms before you reuse it.
Where it fits on the map
Similar projects
Of the same kind, for a shared jurisdiction or tax area. Not a ranking.
- Lemone mBERT taxmBERT model adapted to French tax language, for clustering and semantic search.
- Lemone-Embed LEmbedding model for search over French tax texts, fine-tuned from multilingual-e5-large.
- AI-Powered-Tax-OptimizationRandom Forest model and Flask API that return one optimised tax figure from nine financial inputs.Inactive since 2024Demo
Maintain this project?
Link to this page from your README, so readers find the facts, the notes and where it fits. The badge is a small SVG served from this site.
[](https://tax-oss.com/projects/steuerllm/)