Cisco Time Series Model 1.0

The Cisco Time Series Model 1.0 is Cisco's 250M-parameter univariate zero-shot forecaster and the GA release of the family. It is built for the same observability-style workloads as the preview but is positioned as the production default: smaller and more efficient, with denser probabilistic coverage.

Architecturally it is a multiresolution decoder-only transformer with 25 transformer layers and 15 quantile outputs. Unlike the preview, which was initialized from TimesFM weights, the GA checkpoint is trained from scratch on observability-domain data; the official card attributes the core corpus to Splunk Observability Cloud metrics with additional public time-series sources, and the model is released under Apache-2.0.

On TSFM.ai reach for the 1.0 GA as the default Cisco choice for infrastructure monitoring and network telemetry forecasting, where its multiresolution context and 15-quantile output suit observability series well. Step up to the 500M Cisco Time Series Model 1.0 Preview only if a workload specifically rewards the larger, deeper checkpoint and you can accept that its preview behavior and API surface may still change.