Multi-Modal Foundation · ecmwf

aifs single 0.2.1

Learns over graph-structured data.

Verify the licence before commercial useThe cc-by-4.0 licence carries terms worth reading before you depend on it.cc-by-4.0

Availability: On demand
Typical latency: 12s
Cost per run: $0.02
Compute tier: S · Micro

Not runnable here yet

We index this model but cannot run it: loads, then fails at inference — see FAILS_AT_INFERENCE.

Everything below is still accurate — the licence, the card and the provenance. Running it needs a loader for this repository, not a bigger machine.

About this model

Learns over graph-structured data. Runs on our servers, which pull ecmwf/aifs-single-0.2.1 and execute it as a graph-ml pipeline. The input and output shapes were derived from the task the repository declares rather than written by hand — check the model card before relying on a result.

Standardized I/O contract

Every model in the Hub speaks the same contract, which is what lets the Router and the agent call any of them without special-casing.

Inputs

graphfile · required
Graph

Outputs

vectorvector
Mean-pooled representation, one vector per input.