LLMKube
Kubernetes operator for self-hosted LLM inference with pluggable runtimes (llama.cpp, vLLM, TGI, Ollama, vllm-swift), multi-GPU sharding, NVIDIA CUDA + Apple Silicon Metal support, and OpenAI-compatible API.
Description from awesome-selfhosted (CC BY-SA 3.0)
Still maintained? Yes: 487 commits in the last 90 days, latest release v0.10.14 (Oct 2026).
At a glance
LLMKube quick facts
- Health score
- 90/100 GOOD
- Stars
- 248
- Last commit
- yesterday (Oct 2026)
- Latest release
- v0.10.14, Oct 2026
- Licence
- Apache-2.0
Top alternatives to LLMKube
In its own words
The project's README opens: "Run models across NVIDIA, Apple Silicon, and AMD from one spec. Your models. Your hardware. Your rules."
Why it ranks here
Health score explained
LLMKube scores 90/100, #12 of 14 in Generative Artificial Intelligence (GenAI): 487 commits in the last 90 days earn 35.0 of 35 points, the latest release (v0.10.14, yesterday) earns 25.0 of 25, 248 stars add 9.6 of 20, an active, unarchived repository keeps 10 and Docker support adds the final 10. That places it in the good band (70 and above). The biggest shortfall is star count, worth 10.4 more points.
Recomputed nightly from the GitHub API. How the score is calculated.
Health over time
- Commit activity 35.0 / 35
487 commits in 90 days (30+ scores full marks)
- Release recency 25.0 / 25
v0.10.14 released yesterday
- Stars biggest shortfall 9.6 / 20
248 stars on GitHub (logarithmic)
- Not archived 10.0 / 10
Repository is active
- Docker support 10.0 / 10
Docker image or Compose file available
Commit activity
- 487
- Last 90 days
- 1,217
- Last 12 months
- yesterday
- Last push
Ranked #12 of 14 tools in Generative Artificial Intelligence (GenAI) by health , between GoModel (#11) and Khoj (#13) .
Details
| Licence | Apache-2.0 |
| Platforms | Go, Docker, K8S |
| Latest release | v0.10.14 · yesterday |
| Stars | 248 |
| Open issues | 106 |
| Docker | Yes |
| Repository | defilantech/LLMKube ↗ |
LLMKube vs its top 5 alternatives
Active tools that share a category with LLMKube (Generative Artificial Intelligence (GenAI)), scored on the same five terms.
| Tool | Health | Stars | Licence | Last commit |
|---|---|---|---|---|
| Ollama | 100 | 185k | MIT | today |
| Open-WebUI | 100 | 157k | BSD-3-Clause | today |
| LibreChat | 99 | 46k | MIT | today |
| LocalAI | 99 | 50k | MIT | today |
| GoModel | 92 | 1.3k | MIT | yesterday |
| LLMKube (this tool) | 90 | 248 | Apache-2.0 | yesterday |
LLMKube: common questions
Is LLMKube still maintained?
LLMKube had 487 commits in the last 90 days and its last push was yesterday; the latest tagged release, v0.10.14, is from yesterday. Repoly recomputes this nightly.
Run itSponsored
Run LLMKube on a VPS from $4/mo
LLMKube ships a Docker image, so any of these providers can run it on a small instance.
- is*hosting NVMe VPS in 20+ locations, hourly billing from $5/mo Deploy ↗
- ProHoster VPS with DDoS protection included from $4/mo Deploy ↗
- GODLIKE Game and app servers on NVMe from $4/mo Deploy ↗
Affiliate links. Repoly may earn a commission, which never affects health scores.