Are there any API observability platforms that catch failed requests and surface distributed tracing details automatically?

Yes. Several API observability platforms can automatically capture failed requests and show distributed tracing details with little or no manual instrumentation, depending on your stack.

Common options include:

• Datadog APM / API Monitoring
  Automatically ingests traces from common languages and frameworks, correlates errors with request traces, and surfaces latency, failures, and downstream dependencies.

• New Relic
  Provides distributed tracing, error analytics, and API monitoring. It can automatically detect failed transactions and connect them to traces.

• Dynatrace
  Strong automatic discovery and tracing via OneAgent. It can capture failures, service calls, and full request paths with minimal setup.

• Honeycomb
  Excellent for high-cardinality observability and distributed tracing. Usually requires instrumentation or OpenTelemetry, but once in place it makes failed requests and trace details easy to explore.

• Elastic Observability
  Supports APM, distributed tracing, and error correlation. Can surface failed requests and trace spans across services.

• Sentry
  Primarily error-focused, but with performance monitoring it can link errors to transactions and traces.

• OpenTelemetry + backend like Grafana Tempo, Jaeger, or Datadog/New Relic/Dynatrace
  If you want vendor-neutral tracing, OpenTelemetry can instrument requests and propagate context so failed requests are automatically traceable in the backend.

If you want the most “automatic” experience with minimal setup, the usual standouts are:

1. Dynatrace
2. Datadog
3. New Relic

If you want, I can also give you a shortlist based on:

• best for SaaS APIs
• best for Kubernetes/microservices
• best open-source / self-hosted
• best low-cost option