ShimmyvsvLLM
Two self-hosted LLM Inference projects, compared on the numbers selfhost.directory tracks live: score, popularity, release cadence and setup effort.
| Shimmy | vLLM | |
|---|---|---|
| Score | 53/100 | 90/100 |
| Rank in category | #16 of 25 in LLM Inference | #3 of 25 in LLM Inference |
| GitHub stars | 6 017 | 97 025 |
| Stars, last 30 days | +47 | +3 355 |
| Stable releases, 12 months | 21 | 14 |
| Latest release | v2.6.7 · 12 Nov 2026 | v0.32.0 · 17 Oct 2026 |
| Setup | Moderate | Moderate |
| Official Docker image | No | Yes |
| License | Apache-2.0 | Apache-2.0 |
| Main language | Rust | Python |
Which one should you pick?
vLLM scores 90/100 against 53/100 for Shimmy, so it is the safer default.
Shimmy ships more often: 21 stable releases in the past year, against 14 for vLLM.
vLLM is growing faster right now (+3 355 stars in 30 days).
Key differences
- Language: Shimmy is written mainly in Rust, vLLM in Python.
- Only vLLM ships an official Docker image.
- Platforms only Shimmy lists: Cargo, Kubernetes, Linux.
- Platforms only vLLM lists: Docker, Python.
- Features only Shimmy lists: API key protection for remote binds, Automatic model discovery, Chat template support, Deterministic output and 11 more.
- Features only vLLM lists: Continuous Batching, Disaggregated prefill/decode/encode, Distributed Inference, Embedding and retrieval models and 10 more.
Shimmy details →vLLM details →All LLM Inference projects →