Alibaba launches Qwen2.5-VL, its new AI language model to take on DeepSeek and ChatGPT: What is it, features and more

Alibaba Qwen 2.5

Soon after DeepSeek’s announcement, another Chinese company – Alibaba – launched its latest LLM (Large Language Model) called Qwen2.5. The new LLM from the company positioned it as a competitor to DeepSeek-V3 and other leading AI models like GPT-4 and Gemini. The company claims the model outperforms OpenAI’s GPT-4o, Meta’s Llama 3.1-405B, and DeepSeek-V3. The launch was announced on the first day of the Lunar New Year.

Qwen 2.5 is Alibaba’s latest artificial intelligence model, designed to compete with DeepSeek-V3 and OpenAI’s GPT-4o. Qwen 2.5 is pre-trained on large-scale multilingual and multimodal data and post-trained on quality data to align with human preferences. It supports natural language understanding, text generation, vision and audio processing, AI tool use, and chatbot role-play. The latest update includes enhancements in structured data comprehension and long-text generation.

Qwen 2.5 consists of dense, decoder-only language models ranging from 0.5 billion to 72 billion parameters, available in base and instruct variants. The model is pre-trained on an 18-trillion-token dataset, supporting multilingual capabilities in 29 languages, including Chinese, English, Spanish, and Arabic. It can process up to 128,000 tokens in context and generate outputs of up to 8,000 tokens.

In addition to language capabilities, Qwen 2.5 improves document parsing, object detection, and video understanding. It enhances accuracy in identifying objects across multiple formats, including JSON,

Additionally, the Qwen 2.5 comes with improved video understanding, taking a step up against the competition’s image understanding through ultra-long video processing and fine-grained video grounding. The model applies dynamic resolution and frame rate training for better temporal understanding, helping it extract event segments efficiently. It also integrates a streamlined vision encoder that improves both training and inference speeds.

DeepSeek-V3’s rise has influenced AI market trends, leading to upgrades from Alibaba, ByteDance, and Tencent. DeepSeek’s open-source model previously triggered an AI price war in China, prompting cost reductions among competitors. Alibaba’s latest model aims to provide an alternative with advanced multimodal functions and optimised inference speeds.