Skip to the content
Know your tech before you buy it. Makers: get listed from $49/mo

Google Cloud Pub/Sub

cloud.google.com

  • Free plan
  • Web
  • API
7.4 Club score Visit cloud.google.com

What Google Cloud Pub/Sub is

Google Cloud Pub/Sub is a cloud messaging service that separates services producing messages from those processing them. It is intended for service-to-service communication, with uses including streaming analytics, data integration pipelines, middleware, and parallel task queues—not direct communication with end-user or IoT clients. Communication typically has latency on the order of 100 milliseconds. Synchronous cross-zone replication and per-message receipt tracking provide at-least-once delivery. The service has no shards or partitions and provisions automatically after quota is set. Message controls include seek and replay, dead letter topics, and attribute-based filtering. Pub/Sub integrates with Google Cloud services such as Dataflow, BigQuery, Cloud Storage, Cloud Run, and Eventarc, as well as third-party tools including Splunk, Datadog, Apache Kafka, and ServiceNow. It supports REST and gRPC APIs, with client libraries for several languages, and does not accept API keys for authentication. The listed maximum message size is 10 MB. The first 10 GB per month is free; beyond that, pay-as-you-go throughput is listed at $40 a month, with usage-based throughput beyond 10 GB and storage at $0.27 per GiB-month.

Who it suits

Pub/Sub suits teams connecting services for messaging, analytics, data integration, middleware, or parallel task queues. It is intended for service-to-service communication rather than direct links to end-user or IoT clients.

Good to know, and what to watch

  • At-least-once delivery with cross-zone replication
  • Supports message replay and dead letter topics
  • Offers REST and gRPC APIs
  • Integrates with Google Cloud and third-party tools
  • Does not support API keys for authentication
  • Not intended for direct end-user or IoT communication
  • Maximum message size is 10 MB

Free usage allowance

Free

Pay-as-you-go throughput

$40/mo

Free monthly tier

Free

Pay-as-you-go

$40/mo

Free plan

Getting started

  1. Open the Google Cloud Pub/Sub website
  2. Start with the free trial or free monthly tier
  3. Set your quota for automatic provisioning
  4. Connect services through REST, gRPC, or a client library
  5. Configure delivery, filtering, replay, and dead letter controls

Limits to check

The free monthly tier covers the first 10 GB of data per month. Pay-as-you-go throughput is listed at $40 a month beyond that allowance, with storage at $0.27 per GiB-month.

The Club's review

Google Cloud Pub/Sub is a managed messaging service for asynchronous communication between software services. It is most compelling for teams building Google Cloud pipelines and integrations; its delivery controls are strong, but direct client messaging and API-key authentication are not supported.

Overview

Pub/Sub separates the services that publish messages from those that process them. That reduces the need for both sides to be available at the same time and suits streaming analytics, data integration, service middleware, and parallel task queues. Communication typically has latency around 100 milliseconds, so it is a better fit for background workflows than interactions that depend on an immediate response.

The cloud service provisions capacity automatically after you set a quota, without shards or partitions to manage. Synchronous cross-zone replication and per-message receipt tracking support at-least-once delivery at any scale. That is a useful reliability baseline, but not an exactly-once guarantee: Dataflow supports exactly-once processing of Pub/Sub streams. Delivery guarantees are configurable.

Key features

  • Recovery and routing: Seek and replay let teams revisit messages, dead-letter topics help isolate failed processing, and attribute-based filtering can route messages selectively. These controls are useful for resilient pipelines, though the maximum message size is 10 MB.
  • Integrations: Google Cloud connections include Dataflow, BigQuery, Cloud Storage, Cloud Functions, Cloud Run, App Engine, Cloud Logging, Cloud Monitoring, Workflows, and Eventarc. Third-party integrations include Splunk, Datadog, Informatica, Confluent Cloud, Apache Kafka, Elastic, Databricks, and ServiceNow. This breadth makes Pub/Sub a natural fit for mixed data and service workflows.
  • Security and access: Fine-grained access controls, end-to-end encryption, and HIPAA compliance support sensitive workloads. Authentication is available through client libraries, the Google Cloud CLI, and REST, but not API keys; teams whose clients depend on API-key authentication will need another approach.
  • Interfaces: Standard REST and gRPC APIs and client libraries for seven languages give developers several ways to connect. Pub/Sub is intended for service-to-service communication, not direct communication with end-user or IoT clients.

Pricing

Pub/Sub has a free plan and usage-based paid pricing. The Free monthly tier costs nothing, billed monthly, and covers the first 10 GB of data per month. The Pay-as-you-go plan is $40 a month, billed usage-based monthly, beyond 10 GB; throughput is priced at $40 per TB, with storage at $0.27 per GiB-month. The 10 GB allowance makes the free tier useful for small workloads or evaluation, while production volumes and retained storage can add cost as usage grows.

A separate throughput breakdown gives a Free usage allowance of free for 10 GiB throughput, then Pay-as-you-go throughput at $40 a month after the first 10 GiB per calendar month for Standard message delivery throughput. The free threshold is therefore tied to monthly usage, not a fixed-capacity deployment. A free trial is also available.

Platforms

Pub/Sub is a cloud deployment available through web and API platforms. Its supported protocols are REST and gRPC; it does not offer a self-hosted deployment option.

Who it's for

Choose Pub/Sub for service-to-service messaging, especially when a workflow needs to connect Google Cloud services or feed analytics and integration pipelines. Automatic provisioning, cross-zone replication, replay, and filtering suit teams that need scalable asynchronous delivery without managing partitions. Look elsewhere for direct end-user or IoT client communication, API-key authentication, messages larger than 10 MB, or workloads that require low-latency synchronous responses.

Pros and cons

Pros

  • Automatic provisioning without shards or partitions reduces the capacity-management burden.
  • Cross-zone replication and per-message receipt tracking support at-least-once delivery at any scale.
  • Replay, dead-letter topics, and attribute filtering provide concrete recovery and routing controls.
  • Google Cloud and third-party integrations connect messaging to analytics, storage, observability, and service tools.

Cons

  • At-least-once delivery is not exactly-once processing; exactly-once support for streams is through Dataflow.
  • The 10 MB maximum message size excludes larger payloads.
  • No API-key authentication and no direct end-user or IoT client communication limit some integration designs.
  • Usage-based throughput and storage charges can rise beyond the monthly free allowance.
  • Typical latency around 100 milliseconds makes it a poor fit for immediate-response interactions.

Alternatives

Consider these alternatives when their deployment model, scope, or free offering better matches the job:

  • Lenses offers a free K2K Community plan with Kafka replication and at-least-once delivery, capped at five topic partitions per replication; choose it when that Kafka-focused scope and cap suit the workload.
  • Buf Schema Registry has a free Community plan with one private repository and unlimited public repositories; choose it when managing schemas is the primary need.
  • Confluent Cloud has a free Basic plan with serverless managed clusters, autoscaling Kafka resources, and more than 80 managed connectors; consider it for a Kafka-oriented managed service.
  • TIBCO Messaging Schema Repository for Apache Kafka offers a free Community Edition capped at 100 production application instances or 1,000 web or mobile instances, without TIBCO Support.
  • Sourcemeta One offers a free self-hosted Community edition with a subset of functionality under Business Source License 1.1, with releases transitioning to AGPL-3.0-or-later after four years.
  • Apache Pulsar is a free option.
  • AxonOps Schema Registry is a free, open-source option.
  • Karapace has a free plan capped at 250 KiB/s throughput, three days' retention, and five topics with two partitions each, with Schema Registry and REST proxy included.

For a broader comparison, browse Schema Registry Software, Message Queue (MQ) Software, and Message Broker Software.

Verdict

Choose Google Cloud Pub/Sub for scalable service-to-service messaging when Google Cloud integrations, replay, and at-least-once delivery matter more than direct client access. Its automatic provisioning and recovery controls make it a strong pipeline backbone; look elsewhere if API keys, larger messages, or immediate-response communication are requirements, and budget for throughput and storage beyond the free tier.

The verdict

Pub/Sub provides scalable service-to-service messaging with replay and filtering controls. Teams should account for its authentication limitation, message-size cap, and usage-based charges beyond the free monthly allowance.

Questions people ask

Is Google Cloud Pub/Sub free?

Yes. The free usage allowance is free and covers 10 GiB throughput. The free monthly tier covers the first 10 GB of data per month.

What does paid usage cost?

Pay-as-you-go throughput is listed at $40 a month after the first 10 GiB per calendar month. Storage is listed at $0.27 per GiB-month.

Which protocols and APIs are supported?

Pub/Sub supports REST and gRPC APIs, along with client libraries for seven languages. Authentication supports client libraries, the Google Cloud CLI, and REST, but not API keys.

What is the maximum message size?

The maximum message size is 10 MB.

What delivery guarantees does it provide?

Synchronous cross-zone replication and per-message receipt tracking provide at-least-once delivery. Dataflow supports reliable exactly-once processing of Pub/Sub streams.

What does Pub/Sub integrate with?

Integrations include Google Cloud services such as Dataflow, BigQuery, Cloud Storage, Cloud Functions, Cloud Run, and Eventarc, plus tools including Splunk, Datadog, Apache Kafka, and ServiceNow.

What the maker publishes

APIs and libraries
Pub/Sub uses standard gRPC and REST APIs and provides client libraries for several languages, with the product page stating seven languages. source
Authentication
Pub/Sub supports client libraries, the Google Cloud CLI, and REST, and does not support API keys as an authentication method. source
Automatic scaling
Pub/Sub has no shards or partitions and uses automatic provisioning after you set your quota. source
Delivery reliability
Synchronous cross-zone message replication and per-message receipt tracking provide at-least-once delivery at any scale. source
Exactly-once processing
Dataflow supports reliable exactly-once processing of Pub/Sub streams. source
Founded
1998 source
Headquarters
Mountain View, California, United States source
Integrations
Google Cloud integrations include Dataflow, BigQuery, Cloud Storage, Cloud Functions, Cloud Run, App Engine, Cloud Logging, Cloud Monitoring, Workflows, and Eventarc. source
Intended users
Pub/Sub is intended for service-to-service communication rather than direct communication with end-user or IoT clients. source
Latency
Pub/Sub communication typically has latencies on the order of 100 milliseconds. source
Message controls
Pub/Sub supports seek and replay, dead letter topics, and attribute-based message filtering. source
Messaging service
Pub/Sub is an asynchronous and scalable messaging service that decouples services producing messages from services processing them. source
Security
Pub/Sub offers fine-grained access controls, end-to-end encryption, and HIPAA compliance. source
Support resources
Google Cloud provides Pub/Sub documentation, community forums, support, release notes, and system status resources. source
Third-party integrations
Listed third-party integrations include Splunk, Datadog, Informatica, Confluent Cloud, Apache Kafka, Elastic, Databricks, and ServiceNow. source
Use cases
Pub/Sub is used for streaming analytics, data integration pipelines, service integration middleware, and parallel task queues. source

Checked against the maker's pages on 1 Oct 2026.