Qwen: Qwen3 4B (free)

Qwen: Qwen3 4B (free)

qwen · Released Apr 30, 2025 Emerging
11.5
Our Score

Performance Profile

Intelligence2.6Technical0Value0Content2.5
Intelligence 2.6/10
Technical 0/10
Content 2.5/10
Value 0/10

Qwen3-4B is a 4 billion parameter dense language model from the Qwen3 series, designed to support both general-purpose and reasoning-intensive tasks. It introduces a dual-mode architecture—thinking and non-thinking—allowing dynamic switching between high-precision logical reasoning and efficient dialogue generation. This makes it well-suited for multi-turn chat, instruction following, and complex agent workflows.

40,960 tokens
Context Window
4B Parameters

Capabilities

Tool Use Function Calling

Architecture

ModalityText → Text
TokenizerQwen3
Instruct Typeqwen3
Parameters4B

Performance Indices

Source: Artificial Analysis

12.5 Intelligence Index

Benchmark Scores

Intelligence

GPQA Diamond 39.8% Graduate-level scientific reasoning
HLE 3.7% Humanity's Last Exam
MMLU Pro 58.6% Multi-task language understanding
MATH 500 84.3% Mathematical problem-solving
AIME 21.3% Competition mathematics
SciCode 16.7% Scientific computing

Technical

LiveCodeBench 23.3% Live coding evaluation

Benchmark data from Artificial Analysis and Hugging Face

How does Qwen: Qwen3 4B (free) stack up?

Compare side-by-side with other emerging models.

Compare Models

Model Information

OpenRouter ID qwen/qwen3-4b:free
Providerqwen
Release Date April 30, 2025
Context Length40,960 tokens
Status Active