Source: https://seerepo.com/compare/apache/airflow/vs/microsoft/agent-lightning
Markdown source: https://seerepo.com/compare/apache/airflow/vs/microsoft/agent-lightning.md
Description: airflow and agent-lightning share 5 technologies: FastAPI, uvicorn, jinja2, Pydantic, and structlog. Both are written in Python.

# airflow vs agent-lightning

airflow and agent-lightning share 5 technologies: FastAPI, uvicorn, jinja2, Pydantic, and structlog. Both are written in Python.

- [apache/airflow](https://seerepo.com/r/apache/airflow): Apache Airflow - A platform to programmatically author, schedule, and monitor workflows
- [microsoft/agent-lightning](https://seerepo.com/r/microsoft/agent-lightning): The absolute trainer to light up AI agents.

## What they share and what sets them apart

- In common: FastAPI, uvicorn, jinja2, Pydantic, structlog
- Only in airflow: Flask, aiosqlite, azure-search-documents, azure-storage-blob, confluent-kafka, ibm-db, influxdb3-python, pgvector, presto-python-client, psycopg, psycopg2-binary, pymongo, python-arango, Redis, snowflake-sqlalchemy, weaviate-client, pyarrow, Celery, oracledb, pyiceberg, DuckDB, pinecone, pymssql, clickhouse-connect, gcloud-aio-storage, gcsfs, Neo4j, pymysql, SQLite, Google, flask-appbuilder, asgiref, AIOHTTP, werkzeug, a2wsgi, Apache, Vite, React, react-router-dom, Rich, flask-sqlalchemy, SQLAlchemy, alembic, sqlglot, azure-identity, flask-session, google-auth, pyspnego, pyjwt, cryptography, flask-login, keyring, google-auth-httplib2, google-auth-oauthlib, azure-ai-projects, pydantic-ai-slim, cohere, nbconvert, aiosmtplib, beautifulsoup4, icalendar, sendgrid, flask-wtf, wtforms, jsonschema, msgspec, marshmallow, azure-mgmt-containerinstance, azure-mgmt-resource, Flower, OpenTelemetry, kubernetes, boto3, psutil, kubernetes-asyncio, python-on-whales, docker, Alibaba Cloud, colorlog, org.apache.logging.log4j:log4j-api, org.apache.logging.log4j:log4j-core, org.slf4j:slf4j-api, next-themes, Chakra UI, React Hook Form, yaml
- Only in agent-lightning: None

## Side by side

| | airflow | agent-lightning |
|---|---|---|
| Size | 2,200,257 lines written, 4,596 code files | 16,024 lines written, 35 code files |
| Who writes it | 0.7% of its commits signed by AI. With Claude Code, GitHub Copilot, Cursor, and Devin | 4.5% of its commits signed by AI. With GitHub Copilot, Claude Code, and Cursor |
| Activity | Last push November 5, 2026, 41,804 commits, 47,376 stars | Last push November 5, 2026, 672 commits, 18,906 stars |
| Tests | 4,949 test files. 53.3% of its code files have a test that imports them | 24 test files. 62.9% of its code files have a test that imports them |

- Size: Lines written by people and agents, without generated ones; code files, without tests.
- Who writes it: A commit counts as AI's when it carries its signature: a Co-Authored-By line, its author name or the merge of its branch.
- Activity: From GitHub and its history, on the day SeeRepo read it.
- Tests: The files that test the code, and the share of code files a test imports.

## Frequently asked questions

### Which is bigger, airflow or agent-lightning?

airflow, with 2,200,257 lines written; agent-lightning has 16,024.

The proof: [apache/airflow's code at commit 322c178](https://github.com/apache/airflow/tree/322c178e05b6621b9577e62e66964c1f4c0b590f)

### Which has more GitHub stars, airflow or agent-lightning?

airflow, with 47,376 stars; agent-lightning has 18,906.

The proof: [apache/airflow on GitHub](https://github.com/apache/airflow)

### Which has more tests, airflow or agent-lightning?

airflow, with 4,949 test files; agent-lightning has 24.

The proof: [apache/airflow's code at commit 322c178](https://github.com/apache/airflow/tree/322c178e05b6621b9577e62e66964c1f4c0b590f)

### Which was updated more recently, airflow or agent-lightning?

Both got their last push on November 5, 2026.

The proof: [apache/airflow's commits on GitHub](https://github.com/apache/airflow/commits)

### Which has more commits signed by AI, airflow or agent-lightning?

agent-lightning: an AI signed 4.5% of its commits, with GitHub Copilot, Claude Code, and Cursor; in airflow, 0.7%.

The proof: [the commits each tool signed in microsoft/agent-lightning](https://seerepo.com/r/microsoft/agent-lightning/ai)

## Sources

- apache/airflow: Read on November 5, 2026, at commit 322c178 (https://github.com/apache/airflow/tree/322c178e05b6621b9577e62e66964c1f4c0b590f)
- microsoft/agent-lightning: Read on November 6, 2026, at commit bc3a9e8 (https://github.com/microsoft/agent-lightning/tree/bc3a9e80417d731744040df5d37117f9cc48005f)

## What is your repo built with?

Any public repo, with no account: https://seerepo.com/r/owner/repo/built, with the owner and name it has on GitHub.