DeepSeek V4.1 Flash achieves up to a 1.5x speed boost on the GB300
According to Artificial Analysis, DeepSeek V4.1 Flash saw its AA-AgentPerf speed on the GB300 NVL72 increase by up to 1.5 times tok/s/agent in one month, with per-agent throughput rising from 596 tok/s to 848 tok/s (a 1.42x improvement).
Continue looking at this topic
OpportunitiesManufacturers are building NAS devices into small-model workstations, but the opportunity lies not in selling hardware—rather, in creating plug-and-play industry-specific all-in-one units for sectors hesitant to move data to the cloud."Data should not be uploaded to the cloud."nicheOpen-source projectdeepseek-ai/deepseek-harness ↗Hotly discussed viewsHow do you find the clearest wallpaper using an Agent, and how can you verify it is the original image? ↗News and AnalysisAI agents generated only hundreds of orders: how should we view DoorDash's pilot? ↗Return to original text
A short introduction is retained here; the full text from the source website is not reposted. For specific wording and subsequent updates, please refer to the original article.
Open AIHOT original