LongText-Bench Text Comprehension Benchmark Dataset
LongText-Bench is a text understanding benchmark dataset released by Tencent in 2025. The related paper results are "X-Omni: Reinforcement Learning Makes Discrete Autoregressive Image Generative Models Great Again", which aims to evaluate the model's ability to accurately understand long Chinese and English texts.
The dataset contains 160 prompts for evaluating long text rendering tasks, covering 8 different scenarios (road signs, labeled objects, printed materials, web pages, slides, posters, headlines, and dialogues).
Dataset features:
- Cross-language coverage
- Text length
- Gradient Design
Build AI with AI
From idea to launch — accelerate your AI development with free AI co-coding, out-of-the-box environment and best price of GPUs.
AI Co-coding
Ready-to-use GPUs
Best Pricing
Hyper Newsletters
Subscribe to our latest updates
We will deliver the latest updates of the week to your inbox at nine o'clock every Monday morning
Powered by MailChimp