difyvsllama-swap
Two self-hosted AI Workflows projects, compared on the numbers selfhost.directory tracks live: score, popularity, release cadence and setup effort.
| dify | llama-swap | |
|---|---|---|
| Score | 90/100 | 72/100 |
| Rank in category | #1 of 7 in AI Workflows | #2 of 7 in AI Workflows |
| GitHub stars | 161 269 | 6 259 |
| Stars, last 30 days | +2 982 | +318 |
| Stable releases, 12 months | 16 | 56 |
| Latest release | 1.19.0 · 4 Nov 2026 | v273 · 15 Nov 2026 |
| Setup | Moderate | Moderate |
| Official Docker image | Yes | Yes |
| License | Other | MIT |
| Main language | Python | Go |
Which one should you pick?
dify scores 90/100 against 72/100 for llama-swap, so it is the safer default.
llama-swap ships more often: 56 stable releases in the past year, against 16 for dify.
dify is growing faster right now (+2 982 stars in 30 days).
Key differences
- License: dify is Other, llama-swap is MIT.
- Language: dify is written mainly in Python, llama-swap in Go.
- Platforms only dify lists: Linux.
- Platforms only llama-swap lists: Docker, Homebrew, Pre-built Binaries, Windows.
dify details →llama-swap details →All AI Workflows projects →