Latest Search
Quote
| Back Zoom + Zoom - | ||||
|
BABA-W's Tongyi Qianwen Unveils Qwen3 Series AI Foundation Models
Recommend 20 Positive 33 Negative 11 |
||||
|
||||
|
BABA-W (09988.HK)(BABA.US) unveiled its Qwen3 series of AI foundation models under Tongyi Qianwen, featuring two Mixture of Experts (MoE) models and six dense models with parameters ranging from 0.6B to 235B. The flagship Qwen3-235B-A22B competes strongly against top models like DeepSeek-R1, OpenAI’s o1 and o3-mini, xAI’s Grok-3, and Google’s Gemini-2.5-Pro, excelling in coding, math, and general capability benchmarks. The compact MoE model Qwen3-30B-A3B, with just 10% of QwQ-32B’s active parameters, outperforms it, while the small Qwen3-4B model matches the performance of Qwen2.5-72B-Instruct. AAStocks Financial News |
||||
Copied to Clipboard
 |
||||
