Flyte
An open-source orchestration platform for durable data, machine learning and AI agent workflows written in Python.
These buttons open the developer's own site, repository or store listing in a new tab. wares.gg does not host downloads.
About Flyte
Flyte runs workflows that you write in plain Python, without a separate DSL or YAML files. Tasks declare the resources, container images and secrets they need, and the platform recovers from infrastructure and code failures.
Flyte 2 can be installed with pip, and a local devbox starts a web UI on your machine. It targets teams building data pipelines, ML training and AI agents that need to scale to production with observability. A migration path from Flyte 1 is documented.
Strengths
- Workflows are written in plain Python
- Recovers from infrastructure and code failures
- Local devbox with a web UI for development
Limitations
- Aimed at data and ML engineers rather than general users
- Flyte 1 users need to migrate to Flyte 2
Details
- Pricing
- FreeFree, with the source published on GitHub.
- License
- Open source, license not stated
- Developer
- The Flyte contributors
- Platforms
- Web, Self-hosted, Command line
- How it runs
- Downloadable app, Self-hosted
- Account
- Not required
- Best suited for
- Data and ML teams orchestrating pipelines and AI agents in Python
- Categories
- Automation, Developer tools
- Last verified
- Added
- Provenance
- Facts checked against the developer's own pages and store listings, 1 sources on file.
Alternatives to Flyte
Compare allSoftware that can replace Flyte for an important use case, and what changes if you switch.
Apache Airflow
A self-hosted platform for writing, scheduling and monitoring workflows defined as Python code.
Apache Airflow also defines workflows in Python under Apache-2.0, but you run the scheduler, workers and web server yourself and it lacks Flyte's ML focus.
Prefect
A workflow orchestration tool for scheduling, retrying and observing Python data and ML pipelines.
Prefect defines flows with plain Python decorators and offers built-in retries, with some features belonging to its paid hosted platform.
Dagster
An open-source orchestrator for building, scheduling and monitoring data pipelines, with a hosted Dagster+ option.
Dagster uses an asset-based model with lineage tracking and a hosted Dagster+ option, though its announced merger with Prefect makes the roadmap uncertain.
Kestra
An open-source declarative orchestration platform for scheduled and event-driven workflows, self-hosted or in the cloud.
Kestra uses declarative workflows with more than 2,000 plugins and event triggers, instead of Python code, with some features on enterprise plans.
Luigi
A Python tool from Spotify for building, scheduling and visualising pipelines of batch jobs.
Luigi chains Python batch jobs with dependency resolution and visualisation, but targets batch work rather than ML and AI agent workflows.
Conductor
An open-source, event-driven workflow engine for durable execution of application and AI agent workflows.
Conductor is an event-driven engine for durable microservice and AI agent workflows, with many storage backends but more deployment effort.
Dagu
A self-hosted workflow orchestrator that runs scripts, SSH commands and containers from declarative YAML files.
Dagu is a single binary running YAML workflows over scripts and containers, lighter than Flyte but without Python-defined ML pipelines.
Windmill
An open-source, self-hostable platform for turning scripts into workflows, internal apps and scheduled jobs.
Windmill turns scripts in over 20 languages into workflows, internal apps and schedules, with some features tied to its enterprise edition.
Similar software
Related functionality, not necessarily a direct replacement.
Trigger.dev
An open-source platform for durable background jobs, scheduled tasks and AI agents written in TypeScript.
Apache NiFi
A visual dataflow system for automating the movement and transformation of data between systems.
Axolotl
A command-line tool for fine-tuning language models from YAML configuration files.