Q

Qwen3 0.6B

Developed by Qwen
Qwen3-0.6B is the latest generation of large language model with a parameter scale of 0.6B in the Tongyi Qianwen series. It supports the switching between thinking and non-thinking modes and has powerful reasoning, instruction-following, and intelligent agent capabilities.
Downloads 497.09k
Release Time : 4/27/2025

Model Overview

A language model with 0.6B parameters based on large-scale training, which performs excellently in reasoning, instruction-following, intelligent agent capabilities, and multilingual support, and supports more than 100 languages.

Model Features

Dual-mode reasoning
Uniquely supports seamless switching between thinking mode (complex reasoning) and non-thinking mode (efficient dialogue) within a single model
Enhanced reasoning ability
Surpasses previous-generation models in mathematics, code generation, and common-sense logical reasoning
Human preference alignment
Provides a more natural interaction experience in creative writing, role-playing, and multi-round conversations
Professional intelligent agent ability
Accurately integrates external tools and maintains leading performance in complex intelligent agent tasks

Model Capabilities

Text generation
Complex logical reasoning
Mathematical calculation
Code generation
Multilingual translation
Tool call
Multi-round conversation

Use Cases

Intelligent assistant
Multi-round dialogue system
Build an intelligent assistant that can understand context and maintain coherent conversations
Provide a more natural and personalized dialogue experience
Education
Mathematics problem-solving tutoring
Analyze mathematical problems step by step through the thinking mode
Help students understand the problem-solving ideas rather than just providing answers
Development tool
Code generation and explanation
Generate runnable code according to requirements and explain the implementation logic
Improve the work efficiency of developers
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase