AI People Wiki

Builders & open source

Zhuohan Li

AI Research Scientist, Meta

  • AI Research Scientist, Meta14
  • Lead maintainer and project lead (RL integration, numerics), vLLM project25

Zhuohan Li co-created vLLM, the open-source engine for serving large language models, and co-leads its development; the project's governance document lists Zhuohan Li among the lead maintainers responsible for its overall direction. Zhuohan Li is the second author of the paper that introduced vLLM and its PagedAttention approach to KV cache memory. Zhuohan Li is an AI Research Scientist at Meta focusing on inference, and was previously a Member of Technical Staff at OpenAI, where Zhuohan Li led infrastructure and inference work for gpt-oss.123

At a glance

Current roles
AI Research Scientist +1 more
Companies
Meta, vLLM project
Sector
Company
Education
UC Berkeley
Work listed
2
Sources cited
8
Profile reviewed

Career

Zhuohan Li completed a PhD in computer science at UC Berkeley from 2019 to 2024, advised by Ion Stoica. During that time Zhuohan Li interned at Microsoft Research Asia, Anyscale, and Google Brain and Google DeepMind.1

From June 2024 to August 2025 Zhuohan Li was a Member of Technical Staff at OpenAI, heading infrastructure and inference work for the open-weight gpt-oss models and building training and inference systems for reasoning models such as OpenAI o1. Since August 2025 Zhuohan Li has been an AI Research Scientist at Meta, focusing on inference.1

Research and open-source work

Zhuohan Li's doctoral research was on systems for training and serving large models. Alpa (2022) automated inter- and intra-operator parallelism for distributed deep learning, and AlpaServe (2023), with Zhuohan Li as first author, used model parallelism for statistical multiplexing when serving models. Zhuohan Li's homepage also lists the Vicuna chatbot and Chatbot Arena among Zhuohan Li's projects, and Zhuohan Li co-authored the 2023 paper on judging language models with MT-Bench and Chatbot Arena.6781

The 2023 PagedAttention paper, with Zhuohan Li as second author, introduced vLLM, a serving engine that manages the KV cache in pages to reduce memory waste. Zhuohan Li co-created vLLM and co-leads its development, and the project's governance documents list Zhuohan Li as a lead maintainer.3251

Work and contributions

  • vLLM Open-source code

    Co-created vLLM and co-leads its development; second author of the PagedAttention paper that introduced the system, and listed as a lead maintainer and project lead in the project's governance and committers documents.1325

  • gpt-oss Open weights

    Led infrastructure and inference work for gpt-oss while a Member of Technical Staff at OpenAI, according to Zhuohan Li's homepage.1

Timeline

  1. Began a PhD in computer science at UC Berkeley advised by Ion Stoica.1
  2. Co-authored Alpa, on automating parallelism for distributed deep learning.6
  3. Posted AlpaServe, on model parallelism for deep learning serving, as first author.7
  4. Co-authored "Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena".8
  5. Posted the PagedAttention paper introducing vLLM as second author.3
  6. Joined OpenAI as a Member of Technical Staff.1
  7. Joined Meta as an AI Research Scientist focusing on inference.1

Other documented roles

Roles the sources document that are not confirmed as current.

Where they work

Suggest a correction or removal

Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.

References

8 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026

  1. Zhuohan Li — homepage (about, projects, experience)Zhuohan Li · Official page · accessed Oct 21, 2026
  2. vLLM governance process (project leads and lead maintainers)vLLM project · Documentation · accessed Oct 21, 2026
  3. Efficient Memory Management for Large Language Model Serving with PagedAttention (arXiv 2309.06180)arXiv · Paper · published Sep 12, 2023 · accessed Oct 21, 2026
  4. zhuohan123 (Zhuohan Li) on GitHubGitHub · Official page · accessed Oct 21, 2026
  5. vLLM committers and areas of expertisevLLM project · Documentation · accessed Oct 21, 2026
  6. Alpa: Automating Inter- and Intra-Operator Parallelism for Distributed Deep Learning (arXiv 2201.12023)arXiv · Paper · published Jan 28, 2022 · accessed Oct 21, 2026
  7. AlpaServe: Statistical Multiplexing with Model Parallelism for Deep Learning Serving (arXiv 2302.11665)arXiv · Paper · published Feb 22, 2023 · accessed Oct 21, 2026
  8. Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena (arXiv 2306.05685)arXiv · Paper · published Jun 9, 2023 · accessed Oct 21, 2026