PicoMQ

Durable streams on object storage.

Visit website ↗

Profile updated

Console’s review

· Reviewed by David Mytton

What we like

Each stream is independent and scales with the object store, so disks or capacity to manage. Benchmarked up to 100 MiB/s per stream. Metadata control plane independent from storage (SQL, either an ordered command log in Postgres or SQLite). The client is HTTP. One binary (Rust).

Tradeoffs

Not designed for very low latency - object storage commit often requires tens of milliseconds.

Is PicoMQ a fit?

Based on Console’s review published on 2026-08-27. Summary updated .

Who it’s for

Developers needing scalable streaming solutions

Why consider it

It allows for independent and scalable streams with object storage

Keep in mind

It may not be suitable for applications requiring very low latency

Console is a free weekly newsletter with short reviews of developer tools, covering strengths and tradeoffs. This review reflects Console’s assessment at publication on 2026-08-27. Features and limitations may have changed since the review date. Read our selection criteria and learn about Console.

Tech stack

Runs on
  • Linux
Integrates with
  • Apache Kafka
  • HTTP API
  • PostgreSQL
  • S3
  • SQLite
Written in
  • Rust

Checked

Source and notes

implementation: rust. Source: https://picomq.com/docs/contribute. Quote: Rust 1.98 or newer (rust-toolchain.toml pins the exact version). The default test suite has no external dependencies. integration: http-api. Source: https://picomq.com/docs/. Quote: Clients create named streams, append records, and read them back over HTTP or the Kafka protocol. integration: kafka. Source: https://picomq.com/docs/. Quote: Clients create named streams, append records, and read them back over HTTP or the Kafka protocol. integration: s3. S3-compatible Source: https://picomq.com/. Quote: PicoMQ is durable, real-time streams over HTTP and Kafka, built on S3-compatible object storage. integration: postgresql. Source: https://picomq.com/docs/. Quote: Cluster metadata is an ordered command log in Postgres, or SQLite for a single node. integration: sqlite. Source: https://picomq.com/docs/. Quote: Cluster metadata is an ordered command log in Postgres, or SQLite for a single node. runs_on: linux. Source: https://picomq.com/docs/contribute. Quote: scripts/e2e.sh builds the images and runs scenarios against Postgres, RustFS, and pico in Docker Compose.

View research source ↗

Coverage may be incomplete. “Written in” describes implementation, not supported project languages.

Discover your next favorite tool

Independent devtool reviews and the latest betas, in your inbox every Thursday. Free.