Property-based testing for API contracts: does Hypothesis catch what your unit tests miss?

We've been running Hypothesis on our REST API serializers and it caught three edge cases our unit suite completely missed (empty nested objects, unicode normalization, and a floating-point precision drift in currency conversion). But the feedback loop is slow — each failing example takes 30+ seconds to reproduce in CI. For teams running property-based tests against API contracts: how do you keep the test suite fast? Do you shard by property type, cap examples per run, or use targeted strategies seeded from production traffic logs? Also curious whether anyone integrates Hypothesis with OpenAPI schema validation directly, or keeps them separate.