aquigpt/open0-2-lite
TEXT GENERATIONConcurrency Cost:1Model Size:2BQuant:BF16Ctx Length:32kLicense:apache-2.0Architecture:Transformer0.0K Open Weights Cold

aquigpt/open0-2-lite is a compact 1.72 billion parameter open weights reasoning model developed by Aqui Solutions, fine-tuned on Qwen3 1.7B. It delivers exceptional performance competitive with larger models, optimized for 8-bit precision and featuring a 40K context window expandable to 128K. This model excels in instruction following, multilingual tasks, and grade school mathematics, making it highly accessible for consumer hardware and edge deployment.

Loading preview...

Aqui-open0-2 Lite: Efficient Reasoning Model

Aqui-open0-2 Lite, developed by Aqui Solutions, is a 1.72 billion parameter open weights reasoning model based on Qwen3 1.7B. It is designed for efficiency and strong performance on consumer hardware and edge devices.

Key Capabilities & Features

  • Compact & Efficient: 1.72B parameters, optimized for 8-bit precision, making it suitable for modest hardware.
  • Extended Context: Features a 40,000-token context window, expandable to 128,000 tokens using YARN scaling.
  • Strong Reasoning: Demonstrates exceptional performance in instruction following (73.4% IFEval), general knowledge (67.5% MMLU), and grade school mathematics (63.2% GSM8K).
  • Multilingual Support: Achieves 70.2% on MGSM, indicating robust capabilities across 23+ languages.
  • Open Weights: Released under the Apache 2.0 license, allowing for broad research and commercial use.

Ideal Use Cases

  • Educational & Learning: Assistance with mathematics, general knowledge, and multilingual learning.
  • Lightweight Development: Code generation, algorithm implementation, and technical documentation.
  • Edge AI Applications: On-device assistance, offline reasoning, and integration into mobile, IoT, and embedded systems.
  • Multilingual Support: Cross-language communication, translation, and content creation.