JurinAI/JJ-1-2B

VISIONPricing:Input $0.32 / Cached $0.064 / Output $1.6Concurrent Unit Cost:1Model Size:2.3BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Oct 12, 2026License:cc-by-nc-4.0Architecture:Transformer Open Weights Featherless Exclusive Cold

JurinAI/JJ-1-2B is a 2.3 billion parameter language model developed by JurinAI, specifically designed for making business decisions from text or JSON inputs. It excels at providing choices, yes/no probabilities, and scores, particularly in Japanese and English business workflows. With a 32768-token context length, it offers robust performance for analytical tasks, achieving 94.63% accuracy on business workflows. This model is optimized for practical application in enterprise environments requiring structured decision-making.

Loading preview...

JurinAI/JJ-1-2B Overview

JurinAI/JJ-1-2B is a 2.3 billion parameter language model developed by JurinAI, specifically engineered to assist with business decision-making from text or JSON inputs. It provides structured outputs such as choices, yes/no probabilities, and scores, making it highly suitable for analytical tasks in enterprise settings. The model supports a substantial context length of 32768 tokens, allowing for processing of extensive inputs.

Key Capabilities

  • Business Decision Support: Generates choices, probabilities, and scores from textual or JSON data, aiding in structured decision processes.
  • High Accuracy in Business Workflows: Achieves 94.63% hard-label accuracy on Japanese enterprise workflows, demonstrating strong performance in its primary domain.
  • Multilingual Support: Evaluated for both English and Japanese, with specific strengths in Japanese business contexts.
  • Flexible Deployment: Supports various deployment options including Transformers, CLI/HTTP, GGUF/llama.cpp, and NVIDIA CUDA, offering versatility for integration.

Good for

  • Automated Business Analysis: Ideal for applications requiring automated interpretation of business documents and data to inform decisions.
  • Enterprise Workflow Automation: Suitable for integrating into systems that need to process and act upon structured or unstructured business information.
  • Japanese and English Business Contexts: Particularly effective for tasks involving business decisions in both Japanese and English-speaking environments.
  • Resource-Efficient Deployment: Available in BF16 and Q8_0 quantized formats, enabling efficient inference on various hardware configurations.

Finetunes
1 models