fableforge-ai/NEXUS-Legal

TEXT GENERATIONConcurrent Unit Cost:1Model Size:1.5BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Jul 5, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

NEXUS-Legal by fableforge-ai is a 1.5 billion parameter domain-specialist language model, fine-tuned from Qwen/Qwen2.5-1.5B-Instruct, specifically optimized for legal analysis and contract review. This lightweight model is designed for specialized legal tasks, demonstrating strong performance in domain knowledge, reasoning, and uncensored legal and general queries. It is part of the FableForge NEXUS family, offering a focused solution for law and compliance applications.

Loading preview...

NEXUS-Legal: A Specialized Legal AI

NEXUS-Legal, developed by fableforge-ai, is a 1.5 billion parameter language model built upon the Qwen/Qwen2.5-1.5B-Instruct base. It is specifically engineered as a domain specialist for legal analysis and contract review, offering a lightweight and uncensored solution for legal professionals.

Key Capabilities & Performance

This model demonstrates impressive performance for its size, achieving a 93% score (139/150) in a 30-prompt evaluation. This benchmark covers domain knowledge, uncensored domain and general queries, reasoning, and tool use. Notably:

  • Domain Knowledge: Scored 46/50 (4.6/5 average).
  • Uncensored (Domain & General): Achieved 24/25 (4.8/5 average) in both categories.
  • Reasoning: Scored 24/25 (4.8/5 average).

NEXUS-Legal is part of the broader FableForge NEXUS family, a collection of six specialized models, each tailored to a specific domain like Coder, Security, Medical, Finance, and Science. This model's focus on law and compliance makes it a powerful tool for targeted legal applications.

When to Use NEXUS-Legal

  • Legal Analysis: Ideal for tasks requiring deep understanding of legal texts and concepts.
  • Contract Review: Specialized in processing and analyzing contractual documents.
  • Domain-Specific Querying: Excels in answering questions related to law and compliance.
  • Resource-Constrained Environments: Its lightweight nature (1.5B parameters) makes it suitable for deployment on less powerful hardware, including via Ollama or GGUF quants.