# Build vLLM (optional)

This guide demonstrates how to add Qualcomm Cloud AI backend support to the
[vLLM](https://github.com/vllm-project/vllm) open-source library,
which simplifies the creation of OpenAI-compatible web endpoints and
provides features like continuous batching and other optimizations for
LLM inference and serving.

- [Build for x86\_64](https://docs.qualcomm.com/doc/80-99100-3/topic/build-x86.html)
- [Build for AArch64](https://docs.qualcomm.com/doc/80-99100-3/topic/build-aarch64.html)

Note

Multiple vLLM versions support is available. Select the vLLM version that aligns with your target model and feature requirements. For more information, see [cross feature support matrix](https://docs.qualcomm.com/doc/80-99100-3/topic/Cross-Feature-Matrix.html#reference-to-cross-feature-support-matrix).

## Running a sample inference

python examples/offline_inference/qaic.py
    Copy to clipboard

## vLLM deployment, supported features and capabilities

The vLLM deployment workflow describes the end-to-end process for serving large language models (LLMs) and vision-language models (VLMs) using vLLM on Qualcomm® Cloud AI accelerators.

The vLLM supported features and capabilities section highlights the key functionalities available when running vLLM with Qualcomm Cloud AI backend support. These include optimized model execution, continuous batching, advanced quantization options, large-context support, and integration with performance-enhancing features such as prefix caching and disaggregated serving. Together, these capabilities enable efficient, flexible, and production-ready inference for modern LLM and VLM workloads on AI 100 accelerators.

Last Published: Aug 25, 2026

[Previous Topic
Supported features](https://docs.qualcomm.com/bundle/publicresource/80-99100-3/topics/features.md) [Next Topic
Triton Inference Server](https://docs.qualcomm.com/bundle/publicresource/80-99100-3/topics/index_Triton-Serving.md)