Home
Tools
Models
App
Prompts
News
Topics
Submit for inclusion
English
English 日本語 한국어 Português español Deutsch Русский Français 繁體中文 简体中文
option
Home Close
Home
Tools
Category Ranking Generate Cover Generate Image Skill
Models
Large Language Model Multimodal Model
App
Prompts
Prompts Image Prompts
News
News Flash News
Topics
Submit for inclusion
Submit App Submit Tool
Create an account Sign In
Change Language English Change Language
Settings
English EN 日本語 JA 한국어 KO Português PT español ES Deutsch DE Русский RU Français FR 繁體中文 ZH-TW 简体中文 ZH-CN
Home
Flash News
Content
JustinLewis
JustinLewis
February 4, 2026

SuperCLUEs 2025 Chinese LLM benchmark report reveals global AI competition trends. Anthropics Claude-Opus-4.5-Reasoning leads overall ranking, followed by Googles Gemini and OpenAIs GPT-5.2. Chinese models like Kimi-K2.5 and Qwen3-Max enter global top ten, with Kimi topping code generation and Qwen3 tying for first in math. Domestic models dominate the open-source segment, showing a unique open-closed source co-development ecosystem strength.

SuperCLUEs 2025 Chinese LLM benchmark report reveals global AI competition trends. Anthropics Claude-Opus-4.5-Reasoning leads overall ranking, followed by Googles Gemini and OpenAIs GPT-5.2. Chinese models like Kimi-K2.5 and Qwen3-Max enter global top ten, with Kimi topping code generation and Qwen3 tying for first in math. Domestic models dominate the open-source segment, showing a unique open-closed source co-development ecosystem strength.
0
0
Share Click to copy the link
Comments (0)
0/300
Submit
Author DouglasMartínez DouglasMartínez September 22, 2026
Alibaba appoints senior AI researcher Liu Dayiheng as head of the Qwen large language model project following internal restructuring. Liu will be prominently featured at Alibaba's upcoming Cloud Conference, underscoring the strategic importance of the Tongyi Qianwen team. This appointment clarifies management structure and reflects Alibaba's commitment to strengthening its core AI business operations and organizational framework.
0
0
Author AlbertMiller AlbertMiller September 22, 2026
Shanghai AI Lab and SJTU unveil NCP-ArchPreview, the first 8.9B base model using Next Concept Prediction. Pre-trained on 5.73T tokens, it uses only 51.3% of the token budget to match OLMo-3-7B's loss, boosting convergence by 1.95 times and computational efficiency by 1.74 times. This latent space approach improves downstream performance, including a 6-point gain in GSM8K. Fine-tuning requires just 17M parameters, outperforming LoRA without forgetting, while speculative decoding sees acceptance lengths increase by up to 7.59% in code tasks, offering a new path for inference acceleration.
0
0
Author RoySmith RoySmith September 22, 2026
Tencent Huan Yuan launched Hy Image 3.5 preview, boosting image quality, layout, and editing by 30%. It supports up to five reference images for 2K output at just 0.15 yuan per successful generation, charging only for final products. Outperforming competitors in blind tests, it is now available via Tencent Cloud API and Tencent apps like Yuanbao, with pilot studies showing a 50% surge in usage and improved production efficiency.
0
0
/live/7531
Author LucasWalker LucasWalker September 22, 2026
Xiaomi open-sourced MiMo-V2.6 Pro and Flash models, claiming the top spot among open-source AI with 46 points on Artificial Analysis index. Trained in under six days using recursive self-improvement, the models feature enhanced 3D spatial reasoning and computer operation capabilities. Pricing remains competitive at a fraction of major overseas rivals, with the Desktop client offering up to 20x speed improvements. Full open-source weights, technical reports, and RL task environments are now available, supporting reproducible Agentic RL research.
0
0
/live/7530
Author LucasNelson LucasNelson September 22, 2026
SpaceX AI launched Grok 4.7 on September 21, boasting reasoning speed twice that of competitors at half the cost. Input tokens are priced at $2 per million and output tokens at $6 per million, significantly enhancing cost-effectiveness through superior speed and lower pricing.
0
0
OR