BOULDER, Colo. - ColoradoDesk -- Stream, the leading provider of scalable in-app communication and moderation APIs, has announced the launch of its Labels API, a new capability that enables developers to classify content in real time and build fully customized moderation workflows.

Unlike traditional moderation systems that automatically flag, remove, or queue content for review, the Labels API returns raw moderation labels, giving teams complete control over how to act on the results. Developers can now decide whether to filter, route, tag, or analyze content based on their own business logic.

A New Approach to Moderation: Signals, Not Actions

The Labels API is a server-side text classification endpoint that analyzes content and returns labels such as spam, harassment, or other policy-defined categories without triggering review queues, actions, or webhooks.

This approach is ideal for teams that want to:

• Build custom moderation pipelines
• Power internal tooling and analytics
• Route content dynamically across systems
• Apply different rules across regions, user segments, or use cases

By separating classification from enforcement, Stream enables a more flexible and scalable moderation architecture.

More on Colorado Desk

• Proposition 123 Equity Funds to Support Affordable Housing in Four Communities Across Colorado
• CCHR Calls Out Psychiatry's Pattern of Resistance to Antidepressant Deprescribing
• Boston Industrial Solutions Introduces New Natron® 310 Hyper White UV Ink for Enhanced Printing Performance
• SSE Announces Robin Carhart-Harris Keynote on Psychedelics, Brain Science, and Consciousness Research
• New analysis reveals second job workers keep just 80p in every pound they earn

Built for Customization and Scale

The Labels API integrates seamlessly with Stream's broader moderation platform and supports:

• Custom policies and provider routing (including LLM-based classifiers, NLP models, and blocklists)
• Stateless or persistent workflows, depending on whether teams need historical data for analytics or investigations
• Real-time synchronous responses, allowing immediate use in product flows

Teams can also define their own moderation taxonomy, tailor classification behavior, and plug results into downstream systems such as dashboards, ML pipelines, or internal review tools.

Complementing Full Moderation Workflows

While Stream's Moderation Check endpoint provides an end-to-end workflow with automated actions like flagging or removal, the Labels API is designed for teams that want maximum flexibility and control.

This makes it particularly valuable for:

• AI-driven applications that require custom safety layers
• Marketplaces and social platforms with nuanced moderation rules
• Enterprises building internal trust & safety infrastructure

Enabling the Next Generation of AI-Powered Apps

More on Colorado Desk

• NRE Health Institute Launches International Study Examining Motivations Behind Non-Sexual Nudity
• Colorado: Food Freedom! Governor Polis Signs Tamale Act Into Law
• Governor Polis Takes Executive Action to Improve Colorado's Competency System to Keep Communities Safe, Convenes Competency Task Force
• Colorado: Governor Polis Signs Bills into Law
• Colorado: Governor Polis Signs Laws Saving Coloradans Money on Homeowners Insurance, Supporting Students, and Expanding Pathways For Natural Medicine

As applications increasingly rely on AI-generated and user-generated content, moderation needs are becoming more complex and context-dependent. The Labels API allows developers to treat moderation as a composable system, rather than a fixed pipeline.

With this launch, Stream continues to expand its vision of providing developer-first infrastructure for building safe, scalable, and engaging real-time experiences.

If you want to learn more about the Labels API, visit: https://getstream.io/moderation/docs/go-golang/content-moderation/labels/