Qwen3 4B (free)

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.

by Qwen

Overview

Quick stats across all benchmark runs.

Score

—

0 benchmarks

Avg Latency

0ms

0 requests

Pricing

Free in / Free out

per 1M tokens

Context

41K

tokens

Alternatives

Models with similar or better quality but different tradeoffs

No alternatives found

Run benchmarks on this model to discover alternatives

Benchmark Performance

How this model performs across different benchmarks

No benchmark data available

Run benchmarks with this model to see performance breakdown

Score Over Time

Performance trends across all benchmark runs

No score trend data

Score history will appear here after multiple runs

Benchmark Activity

Number of benchmark runs over time

No activity data

Activity will appear here after benchmark runs

Quickstart

View on OpenRouter

Get started with this model using OpenRouter

import { OpenRouter } from "@openrouter/sdk";

const openrouter = new OpenRouter({
  apiKey: "<OPENROUTER_API_KEY>"
});

const completion = await openrouter.chat.completions.create({
  model: "qwen/qwen3-4b:free",
  messages: [
    {
      role: "user",
      content: "Hello!"
    }
  ]
});

console.log(completion.choices[0].message.content);

Get your API key at openrouter.ai/keys

Other Models from Qwen

Compare performance with other models from the same creator

Model	Latency	Cost/1M	Score
Qwen3.6 Max Preview	21.5s	$3.64	53%
Qwen3.6 Plus	33.1s	$1.14	45%
Qwen3.6 Flash	9.6s	$0.66	38%
Qwen3 235B A22B Instruct 2507	29.7s	$0.09	—
Qwen3 VL 8B Instruct	123.8s	$0.29	—
Qwen3 Next 80B A3B Instruct	23.1s	$0.59	—
Qwen2.5 VL 72B Instruct	22.1s	$0.50	—
Qwen2.5 Coder 32B Instruct	19.7s	$0.83	—
Qwen3 Coder Flash	13.3s	$0.58	—
Qwen3 Coder 30B A3B Instruct	30.8s	$0.17	—
Qwen2.5 7B Instruct	12.6s	$0.07	—
Qwen3 Coder 480B A35B	17.1s	$1.01	—
Qwen3.7 Plus	—	$1.00	—
Qwen3 235B A22B	78.3s	$1.14	—
Qwen3 VL 235B A22B Instruct	37.5s	$0.54	—
Qwen3 32B	28.3s	$0.18	—
Qwen3 8B	111.5s	$0.22	—
Qwen3 30B A3B Thinking 2507	44.0s	$0.24	—
Qwen3.5 Plus 2026-04-20	—	$1.05	—
Qwen3 VL 30B A3B Instruct	38.3s	$0.33	—
Qwen3.7 Max	—	$2.50	—
Qwen3 VL 235B A22B Thinking	112.0s	$1.43	—
Qwen3 Coder Plus	20.5s	$1.95	—
Qwen3 235B A22B Thinking 2507	248.4s	$0.10	—
Qwen3 Next 80B A3B Thinking	28.8s	$0.44	—
Qwen Plus 0728	19.6s	$0.52	—
Qwen3 VL 30B A3B Thinking	83.6s	$0.85	—
Qwen3 30B A3B Instruct 2507	21.6s	$0.12	—
Qwen3.5-122B-A10B	—	$1.17	—
Qwen3.5-Flash	14.7s	$0.16	—
Qwen-Plus	22.0s	$0.52	—
Qwen3 VL 32B Instruct	21.0s	$0.26	—
Qwen3 Max Thinking	—	$2.34	—
Qwen3 30B A3B	149.8s	$0.27	—
Qwen2.5 72B Instruct	22.2s	$0.38	—
Qwen3 Max	31.7s	$2.34	—
Qwen3 14B	70.4s	$0.17	—
Qwen3 VL 8B Thinking	79.4s	$0.74	—
Qwen3 Coder Next	2.0s	$0.45	—
Qwen2.5-VL 7B Instruct	10.9s	$0.20	—
Qwen2.5 Coder 7B Instruct	6.4s	$0.06	—
Qwen2.5 VL 32B Instruct	39.1s	$0.14	—
QwQ 32B	143.6s	$0.28	—
Qwen VL Plus	10.4s	$0.42	—
Qwen VL Max	36.4s	$2.00	—
Qwen-Turbo	18.6s	$0.13	—
Qwen-Max	12.7s	$4.00	—
Qwen3.5 Plus 2026-02-15	44.8s	$0.91	—
Qwen3.5-27B	55.5s	$0.88	—
Qwen3.5-9B	112.8s	$0.09	—
Qwen3.5 397B A17B	3.5s	$1.36	—
Qwen3.6 35B A3B	13.4s	$0.57	—
Qwen3.5-35B-A3B	43.1s	$0.57	—
Qwen3.6 27B	22.3s	$1.74	—