DeepSeek's Zhihu post: V4.1 Flash cluster serves 155T tokens/day at $38 per GPU-hour
Swyx cites DeepSeek's Zhihu post claiming a single V4.1 Flash cluster serves 155 trillion tokens per day and earns $38 per GPU-hour at off-peak prices, against $5-10 GPU cost.
deepseek posted way more alpha on their zhihu page, if you only read twitter you are reading the laowai approved stuff "Based on this framework, we have implemented a high-throughput, low-latency large-scale inference service for DeepSeek V4.1 Flash. Currently, the service throughput of a single cluster is 155 trillion tokens/day, and the overall revenue is $38 per card per hour (based on off-peak prices)." it costs roughtly $5-10 per gpu hour so depseek is claiming 80% serving margin at THEIR prices lololololol
This post quotes another post: “DeepSeek open-sources DeepInfer, the inference framework behind V4.1 Flash”