SGLang

SGLang is an inference framework hosted by LMSYS for serving large language, vision-language and diffusion models, from one accelerator to distributed clusters. It is Apache 2.0 licensed: install it with pip or Docker and query models through OpenAI-compatible endpoints.

• Serves large language, vision-language and diffusion models
• OpenAI-compatible endpoints for querying the models you launch
• Runs on NVIDIA and AMD GPUs, Google TPUs, Intel GPUs and CPUs, Apple Silicon and Huawei Ascend NPUs
• Disaggregated prefill/decode, speculative decoding, parallelism and a zero-overhead scheduler
• Built-in SGLang Diffusion engine for image and video generation
• Install with uv or pip, or pull the Docker image

Tasks

Host & deploy apps, Use LLM & AI model APIs

People

ML engineers serving open models on their own GPUs, Teams running large-scale LLM inference, Researchers generating RL rollouts

• You provide the GPUs or other hardware it runs on
• Set up from the command line, with Docker or a Python environment

Is SGLang free?

Yes. SGLang has a free plan: Open source (Apache 2.0): free to self-host on your own hardware. We last checked github.com's pricing page on 2026-10-24.

What does the SGLang free plan include?

Open source (Apache 2.0): free to self-host on your own hardware. Source: https://github.com/sgl-project/sglang (checked 2026-10-24).

How do I get SGLang for free?

Steps: Pull the Docker image or install sglang with uv pip; Launch the server pointing to your model; Query it through the OpenAI-compatible API. Start at https://github.com/sgl-project/sglang (checked 2026-10-24).

Do I need an account to use SGLang for free?

Not for the free option: the free route is to self-host SGLang (Pull the Docker image or install sglang with uv pip). Checked 2026-10-24.

What's the catch with SGLang's free plan?

Runs on your own GPUs, TPUs, NPUs, CPUs or Apple Silicon. Checked 2026-10-24.

What are free alternatives to SGLang?

Exa (free credits), Requesty (free plan) and Crawl4AI (free plan) are free alternatives. We checked their free offers on 2026-10-25.

What hardware does SGLang run on?

NVIDIA and AMD GPUs, Google TPUs, Intel GPUs and Xeon CPUs, Apple Silicon Macs, Huawei Ascend NPUs and Moore Threads GPUs, with more integrations in progress.Source: github.com/sgl-project/sglang

Can SGLang generate images and video?

Yes. SGLang Diffusion, its built-in image and video generation engine, is included in the repository and the sglang Python package.Source: github.com/sgl-project/sglang

How do I send requests to a model served by SGLang?

Launch the server with a command pointing to your model, then use its standard OpenAI-compatible endpoints.Source: sglang.io

• SGLang vs vLLM · free plans side by side

Free offer

Verified 2026-10-24 on github.com/sgl-project/sglang/blob/main/LICENSE

…is licensed under the Apache License…

Page content

Checked 2026-10-24 against

• github.com/sgl-project/sglang
• sglang.io
• github.com/sgl-project/sglang/blob/main/LICENSE

How we check free offers