vLLM
Inference & Compute · Free + paid
High-throughput and memory-efficient LLM serving
At a glance
- Made by
- vLLM community project (started at UC Berkeley Sky Computing Lab)
- Korean
- None (English and a few others only)
- API
- Yes
- Available on
- API, CLI
- Open source
- Apache-2.0
- GitHub
- 93,741 stars · Apache-2.0
What is vLLM?
vLLM is a fast and easy-to-use LLM inference and serving library, a community open-source project that started at the UC Berkeley Sky Computing Lab and is maintained by more than 2,000 contributors. It improves GPU utilization with PagedAttention and continuous batching, and connects directly through an OpenAI-compatible API. It supports more than 200 Hugging Face model architectures on a wide range of hardware, including NVIDIA, AMD, and Intel GPUs, x86/ARM/PowerPC CPUs, and TPUs. Install it with uv or pip.
- Attention KV memory management based on PagedAttention
- Continuous batching, chunked prefill, and prefix caching
- Quantization including FP8, INT4, GPTQ/AWQ, and GGUF
- Speculative decoding such as n-gram and EAGLE
- Distributed inference with tensor, pipeline, data, expert, and context parallelism
- OpenAI-compatible API server, Anthropic Messages API, and gRPC
- Multi-LoRA for dense and MoE models
- Wide hardware support including NVIDIA, AMD, and Intel GPUs, CPUs, and TPUs
Who uses vLLM and for what
- Serving open-source LLMs on a variety of hardware
- Connecting self-hosted models to existing apps through the OpenAI-compatible API
Pricing
Open source (free)
커뮤니티 오픈소스 LLM 서빙 라이브러리(uv·pip 설치, 후원금으로 운영), 요금 표기 없음
공식 페이지 확인 2026-10-14(산출 규칙 2026-10-15)
AI tools similar to vLLM (alternatives)
vLLM vs. alternatives at a glance
| AI | Pricing | Combos on the map | Connected tools | Free | Cheapest monthly | Korean |
|---|---|---|---|---|---|---|
| vLLM | Free + paid | 0 | 5 | — | — | No |
| llama.cpp | Free + paid | 0 | 6 | — | — | — |
| ONNX Runtime | Free + paid | 0 | 6 | — | — | No |
| SGLang | Free + paid | 0 | 6 | — | — | No |
| Sol Engine | Free + paid | 0 | 6 | — | — | No |
0 of these 5 tools have a free plan.
Combo and connection counts are based on the relationships recorded on AI Galaxy Map (2026-10-26); they are not a quality ranking.
vLLM on AI Galaxy Map
- One of 43 AI tools in the Inference & Compute field.
- Connected to 5 other tools — #16 in its field.
Frequently asked questions
Can I use vLLM for free?
Open source (free). Prices can change, so check the official site before paying.
What AI can I use instead of vLLM?
Tools that do the same job, such as llama.cpp, ONNX Runtime, SGLang, Sol Engine, are connected to it on the map.
Can I use vLLM in Korean?
None (English and a few others only)
Sources (official pages)
Checked on the official pages on 2026-10-14. Prices and features can change, so check the official site again before paying.