Product updates, practical guides, and perspectives from PicoX.
A developer's guide to evaluating Kimi K3.1 Pro: a brief look at the announcement and official scores, followed by practical ways to check artifacts, organize context, and measure time and cost per accepted task.
Read articleA closer look at the 1017 update: 11 official benchmarks, coding and agent improvements, a million-token context, peak and off-peak costs, and a practical PicoX API example.
Read articleA fast model needs a well-organized inference system. A closer look at DeepInfer, from prefill and decode to asynchronous scheduling, DSpark, and the conditions behind its performance charts.
Read article
PicoX