Over the past five months since the release of Qwen2-VL, developers have built new models based on it, contributing valuable feedback. Now, Qwen2.5-VL introduces enhanced capabilities, including precise analysis of images, text, and charts, as well as object localization with structured JSON outputs. It understands long videos, identifies key events, and functions as an agent, interacting with tools on computers and phones. The model's architecture features dynamic video processing and an optimized ViT encoder for improved speed and accuracy.
| Claude Sonnet 4 | Qwen2.5-VL-32B | |
|---|---|---|
Provider | ||
Web Site | ||
Release Date | May 22, 2025 3 days ago | Mar 25, 2025 2 months ago |
Modalities | text images | text images video |
API Providers | Anthropic API, Amazon Bedrock, Google Cloud's Vertex AI | - |
Knowledge Cut-off Date | Unknown | Unknown |
Open Source | No | Yes (Source) |
Pricing Input | $3 per million tokens | $0 |
Pricing Output | $15 per million tokens | $0 |
MMLU | 86.5% Source | 78.4% Source |
MMLU Pro | - | 49.5% |
MMMU | 74.4% Source | 70% |
HellaSwag | - | Not available |
HumanEval | - | Not available |
MATH | - | 82.2% |
GPQA | 75.4% Diamond Source | 46.0% Diamond |
IFEval | - | Not available |
Array | - | - |
AIME 2024 | - | - |
AIME 2025 | 75.5% Source | - |
Array | - | - |
Array | - | - |
Array | - | - |
Array | - | - |
Mobile Application | - |
Compare AI. Test. Benchmarks. Mobile Apps Chatbots, Sketch
Copyright © 2025 All Right Reserved.