GO
| HSI1 | 23,093.37 | +158.69 | 249.78B |
| HSCEI1 | 7,756.25 | +67.06 | 87.14B |
| Back Zoom + Zoom - Block Traded | |
|
2026-10-22 10:36:35 Morgan Stanley issued a report stating that MINIMAX-W (00100.HK) officially launched M3.1 Flash on Monday (19th), with parameters and licensing details yet to be disclosed. The model demonstrated significant improvements in agent workflows, programming, computer usage and cybersecurity. It scored 73.7% on DeepSWE, 61.7 on OSWorld V2 and 82% on CyberGym, with the latter two benchmarks ranking among the top Chinese models. The broker noted that although pricing is not the lowest, it remains competitive among China's major flash large language models, while agent task costs also hold an advantage. Optimized sparse attention improved inference speed by nearly 3x and reduced agent task costs by 70%. API pricing is set at RMB1 per million input tokens and RMB4 per million output tokens. M3.1 Flash ranks second among China's major flash LLMs in the Artificial Analysis intelligence index, behind only DeepSeek V4.1 Flash Oct. The combination of stronger intelligence, higher throughput and lower costs should help expand enterprise adoption. The broker expects M3 Pro to become the primary large language model showcasing MiniMax's pursuit of frontier intelligence capabilities, and anticipates its launch before the end of Oct. The broker maintained its Overweight rating and HKD900 TP on MINIMAX-W. AASTOCKS Financial News Website: www.aastocks.com This article was automatically translated by AI, the Chinese version should be considered the authoritative version. AASTOCKS.com Limited does not guarantee its accuracy or completeness and accepts no liability for any damages or losses arising from the use of this translation. | |