# Streaming

**Canonical:** https://apis.io/providers/streaming/  
**APIs profiled:** 21

Streaming is a topic catalog of the protocols, platforms, and processing engines used to move and transform real-time, high-volume, often bidirectional data over the network. It indexes the canonical log-structured and broker systems (Apache Kafka, Apache Pulsar, Redpanda, NATS JetStream, AWS Kinesis, GCP Pub/Sub + Dataflow, Azure Event Hubs, Confluent Cloud, StreamNative), the over-the-wire streaming protocols exposed to API consumers (Server-Sent Events, WebSocket, gRPC streaming, GraphQL subscriptions), the change-data capture and connector frameworks that feed them (Kafka Connect, Debezium), and the stream-processing engines that consume them (Apache Flink, Spark Structured Streaming, Materialize, Tinybird, Bytewax, Apache Beam). This topic is distinguished from `events` and `async-apis`: streaming emphasizes real-time, high-throughput, partitioned, and often bidirectional pipes, rather than discrete event envelopes or static contract documents.

## Kin Score — 14.8 / 100 (emerging)

Scored 2026-08-25 under rubric 0.14.0. Trend: flat (+0.0 from 14.8).

| Facet | Score |
|---|---|
| Discoverability | 64.8 |
| Contract Quality | 21.3 |
| Governance | 25.0 |
| Contract Governance | 25.0 |
| Operational Transparency | 0.0 |
| Developer Ergonomics | 0.0 |
| Commercial Clarity | 0.0 |
| Access Clarity | 0.0 |

## Agent readiness — 1.3 (human-only)

| Dimension | Value |
|---|---|
| Spec Presence | no |
| Agentic Access | no |
| Reversibility Documented | no |
| MCP Server | no |
| Auth Clarity | no |
| Idempotency | no |
| Error Semantics | no |
| OpenAPI Examples | documented |
| Rate Limit Signal | no |
| Event Surface Described | no |
| Agent Skills | no |
| Well Known Catalog | no |
| Consent Identity | no |
| Agent Card | no |
| Dry Run Mode | no |
| Delegated Identity | no |
| Protected Resource Metadata | no |
| Dynamic Client Registration | no |
| Agentic Commerce | no |

## Access

Unknown — onboarding: unknown, pricing: unknown, trial: no (confidence: low).

## APIs (21)

- **Apache Kafka** — Distributed, partitioned, replicated log. The reference open-source streaming platform; durable, ordered topics with consumer groups, exactly -once semantics, and the de facto w...
- **Apache Pulsar** — Cloud-native, multi-tenant pub/sub and streaming platform with a tiered storage architecture (BookKeeper) that separates compute from storage, native geo-replication, and built-...
- **Redpanda** — Kafka-API-compatible streaming platform implemented in C++ with no ZooKeeper/JVM dependency. Single binary, thread-per-core architecture, Raft consensus; positioned as a drop-in...
- **NATS JetStream** — Persistence layer for the NATS messaging system providing at-least-once and exactly-once streaming, key/value and object stores, and durable consumers — designed for edge, IoT, ...
- **Amazon Kinesis** — AWS managed family for real-time streaming: Kinesis Data Streams (shards, partition keys, 24h–365d retention), Kinesis Data Firehose (delivery to S3/Redshift/OpenSearch), and Ki...
- **Google Cloud Pub/Sub and Dataflow** — GCP's managed messaging (Pub/Sub) and stream-processing (Dataflow, built on Apache Beam) stack. Pub/Sub provides at-least-once and exactly-once delivery with push/pull subscribe...
- **Azure Event Hubs** — Microsoft's managed big-data streaming platform; Kafka-protocol compatible, partitioned, with Capture (delivery to ADLS/Blob) and tight integration with Azure Stream Analytics a...
- **Confluent Cloud** — Managed Kafka by the original Kafka authors. Cluster, topic, connector, KSQL, Schema Registry, Stream Governance, and Flink offerings exposed via a Confluent Cloud REST API and ...
- **StreamNative** — Managed Apache Pulsar as a service from Pulsar's original contributors, with multi-cloud clusters, Functions, sources/sinks, and a control-plane REST API.
- **Server-Sent Events (SSE)** — One-directional HTTP-based streaming from server to client using the `text/event-stream` media type. Defined by the HTML Living Standard EventSource API; widely used for LLM tok...
- **WebSocket** — Full-duplex, bidirectional streaming protocol over a single TCP connection, upgraded from HTTP. RFC 6455. Foundation for chat, collaborative apps, market data, and real-time con...
- **gRPC Streaming** — gRPC defines four RPC styles, three of which are streaming: server streaming, client streaming, and bidirectional streaming, all multiplexed over HTTP/2. The default streaming s...
- **GraphQL Subscriptions** — The GraphQL operation type for receiving a stream of updates over a long-lived transport (typically WebSocket via the graphql-ws or graphql-transport-ws sub-protocols, or SSE). ...
- **Kafka Connect** — Framework and runtime for source/sink connectors that move data into and out of Kafka. Distributed mode runs a REST-controlled cluster of workers managing connector and task lif...
- **Debezium** — Change-data-capture (CDC) platform that streams row-level database changes (Postgres, MySQL, MongoDB, SQL Server, Oracle, Cassandra) as Kafka records using each database's nativ...
- **Apache Flink** — Distributed, stateful stream-processing engine with event-time semantics, windowing, watermarks, and exactly-once state. SQL, DataStream, and Table APIs; reference engine for su...
- **Spark Structured Streaming** — Stream-processing API built on Spark SQL using a micro-batch (and experimental continuous) execution model. Treats a stream as an unbounded table.
- **Materialize** — Operational data warehouse and streaming SQL database built on Differential Dataflow. Maintains incrementally updated materialized views over streaming sources with millisecond ...
- **Tinybird** — Real-time analytics platform built on ClickHouse; ingests streams via HTTP, Kafka, or CDC, exposes SQL pipes as parameterized HTTP API endpoints with auth tokens.
- **Bytewax** — Open-source Python-native stream-processing framework built on Timely Dataflow; targets data scientists and Python teams building real-time ML and data pipelines.
- **Apache Beam** — Unified batch and streaming programming model. Beam pipelines run on multiple runners (Dataflow, Flink, Spark, Samza), defining the canonical event-time / watermark / window / t...

## Security (2)

- **Streaming Domain Security** — TLSv1.3 · HSTS · DNSSEC · DMARC
- **Streaming Vulnerability Disclosure** — security.txt · contact published

## Tags

Streaming, Real-Time, Event Streaming, Change Data Capture, Stream Processing, Server-Sent Events, WebSocket, gRPC, GraphQL Subscriptions, Kafka, Pulsar, Kinesis, Flink

---

Profiled by [API Evangelist](https://apievangelist.com) and published on [APIs.io](https://apis.io/providers/streaming/). Scores are computed from the provider's own public artifacts under a published rubric.