---
title: "Qwen 3 32B Benchmarks - RankedAGI"
description: "Detailed benchmark and metadata record for Qwen 3."
source: "https://rankedagi.com/models/qwen-3-32b"
---

# Qwen 3

| Field | Value |
| --- | --- |
| Organization | Alibaba |
| License | Open Source |
| Size | 32B |
| Version | Latest |
| Released | 2025-04-29 |
| Context window | 128K |
| Last updated | 2026-04-20T06:24:00.465Z |

## Benchmarks

| Benchmark | Category | Value | Description | Source |
| --- | --- | --- | --- | --- |
| RankedAGI Coding | coding | 45.9% | RankedAGI Coding Score |  |
| RankedAGI Agentic | Agents | 26.9% | RankedAGI Agentic Score |  |
| LiveCodeBench v5 | coding | 65.7% | LiveCodeBench - Evaluation of Large Language Models for Code |  |
| Codeforces ELO Score | coding | 1977 | Codeforces ELO Score |  |
| Aider Polyglot | coding | 40.0% | Aider Polyglot Code Completion Benchmark |  |
| RankedAGI Reasoning | reasoning | 46.3% | RankedAGI Reasoning Score |  |
| AIME 2025 | math | 85.7% | AIME 2025 Competition Math |  |
| AIME 2024 | math | 81.4% | AIME 2024 Competition Math |  |
| Hughes Hallucination Evaluation | safety | 2.8% | Computed using Vectara's Hallucination Evaluation Model, also known as HHEM. This evaluates how often an LLM introduces hallucinations when summarizing a document. | https://github.com/vectara/hallucination-leaderboard |
| RankedAGI Math | math | 68.0% | RankedAGI Math Score |  |
| RankedAGI Overall | general | 44.5% | Overall RankedAGI score |  |
| GDPval-AA Elo | Agents | 493 | Office Tasks (Artificial Analysis) |  |

## Model Links

| Title | URL |
| --- | --- |
| Release | https://qwenlm.github.io/blog/qwen3/ |