Open lightweight ai model excels at multiple Japanese legal tasks
XLaw-JP-27B: A Lightweight Open Model Specialized for Japanese Law
Computation and Language
Summary
Legal questions in Japanese law can be very challenging, especially for AI models due to unique language and legal details. The authors introduce XLaw-JP-27B, a specialized AI model that handles Japanese legal exams and tasks well, running on common hardware with open access. It combines legal text retrieval with smart tuning techniques and performs better than other open models. Its design focuses on accuracy in bar exams and contract review without using future test data.
What this means in practice
- •For legal technology developers: Build AI-powered tools that answer Japanese legal exam questions and support legal research with a lightweight, open model fine-tuned on relevant data.
- •For law firm technology teams: Automate contract drafting and review using the specialized AI model trained on government contract templates tailored for Japanese law.
Authors
Andrew Shin
Abstract
We present XLaw-JP-27B, an open-weight model specialized for Japanese law that runs on a single 48\,GB GPU under the Apache~2.0 licence. It adapts Qwen3.8-27B with hybrid retrieval over all current statutes and 31k case summaries, a light QLoRA fine-tune on its own verified answers to past bar-exam questions (self-distillation), a contract-specific LoRA adapter, and four-sample voting. We evaluate it on eight Japanese legal tasks---the multiple-choice section of the bar examination in 2024, 2025 and 2026, statement-level bar questions (JBE-QA), statute questions (LawQA-JP), the patent attorney exam, and contract review and drafting built from government model contracts---against open models of 12--33B under identical prompts and retrieval. XLaw-JP-27B ranks first on six tasks, ties on one and is third on one (mean rank 1.3 against 2.8--3.1 for the baselines). It is the only system that passes the bar exam in all three years (129, 120 and 130 of 175, meeting every subject minimum; no test-year material was used in training), while open Japanese 32--33B models fail at least one year. A controlled study of each component shows that retrieval, voting and self-distillation help, and that thinking mode, retrieval-grounded synthetic tuning, self-verification, per-statement decomposition and more self-distillation do not. The largest remaining gaps to frontier models are in contract review and patent law. The model is available as a free web service at https://x-law.jp/ with the weights available at https://huggingface.co/X-LawJapan/xlaw-jp-27b under Apache~2.0.