OptiSift Workspace

AI Model Rankings

Compare large language models on price, context window, capabilities, and real benchmark scores.

Private workspace

APEX Observatory

Workspace health

Read-only operational intelligence for coverage, source quality, freshness, duplicate risk, and provenance readiness.

Deterministic snapshot · no remediation actions

Metric coverage

80%

122 of 153 expected metric cells populated.

Healthy

Source reliability

75%

3 of 4 sources meet the current reliability threshold.

Healthy

Freshness evidence

100%

4 of 4 sources include a last-verified timestamp.

Healthy

Provenance readiness

100%

4 sources include both a retrievable URL and verification date.

Healthy

Duplicate candidates

0

Normalized manufacturer and item-name collisions requiring human review.

Healthy

Conflict detection

Pending

Value-level source assertions are not yet exposed in the landing dataset, so conflicts are not inferred.

Instrumentation pending

Current interpretation

This first Observatory increment reports only evidence available in the current workspace dataset. It does not manufacture confidence, freshness, or conflict signals where value-level provenance is absent. The next data lane will add assertion-level provenance, duplicate review states, and explicit conflict records.