NeoSyntropy

MODELS / INTELLIGENCE PRIMITIVES

AI Runtime /
Atoms

Three specialized responsibilities. Combine routing, reasoning, and structured extraction inside the paths your runtime defines.

01 / AI RUNTIME ATOM

Router

Select the next action from the candidates your runtime permits.

Explore Router β†’
02 / AI RUNTIME ATOM

Reason

Work through bounded reasoning steps with verification before progression.

Explore Reason β†’
03 / AI RUNTIME ATOM

Schema

Extract structured fields that match the schema your application defines.

Explore Schema β†’

APACHE-2.0 CHECKPOINTS / 14 MODELS

Models we support

This catalog includes only checkpoints listed with an Apache-2.0 license. Each card links to the publisher's model README and shows the result of our single-L4 smoke test. Passing inference checks are not a quality benchmark.

01 / ROUTING

Router models

Score the legal destinations supplied by your graph. The router ranks candidates; the runtime checks that the selected transition is permitted.

02 / STRUCTURED EXTRACTION

Schema models

Extract entities, fields, or labels for validation against the schema your application defines.

03 / REASONING

Reasoning models

The same checkpoints support both reasoning patterns: stochastic nodes let the model choose among permitted paths; deterministic nodes run developer-authored steps in a fixed order.

β€œDeterministic” describes the workflow order, not a guarantee that generated text or tool choices are identical on every run.