# lm-sys/FastChat

An open platform for training, serving, and evaluating large language models. Release repo for Vicuna and Chatbot Arena.

Repository: https://github.com/lm-sys/FastChat
Canonical: https://ross.abutalabs.com/products/fastchat
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-05-01T00:25:53+00:00

## Health v2 (maintenance only)
Score: 57/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 80, release rhythm 8, longevity 90
- inputs: {"age_days": 1264, "days_push": 125, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 39525, forks 4781 (observed 2026-08-28T04:12:07.893004+00:00)

## What it is
FastChat is an open platform for training, serving, and evaluating large language model based chatbots, originally releasing the Vicuna model and powering Chatbot Arena. It provides a distributed multi-model serving system with a web UI and OpenAI-compatible RESTful APIs, plus training and evaluation code such as MT-Bench.

## Use cases
- serve local LLMs behind an OpenAI-compatible API
- host a chatbot web UI for multiple models
- fine-tune chat models like Vicuna
- evaluate chatbots with MT-Bench
- run side-by-side LLM comparisons and collect human votes
- build an LLM leaderboard from human preference data

## When to choose
- you need a distributed multi-model LLM serving system with OpenAI-compatible endpoints
- you want to train or fine-tune open chat models like Vicuna
- you need benchmarking and human evaluation tooling for chatbots

## When to avoid
- you only need lightweight single-model inference without a serving platform
- you need the newest serving features, as development has slowed and newer stacks like vLLM may be better maintained
- you want a managed cloud service rather than self-hosted infrastructure

## Facets
- artifact type: framework
- maturity: maintenance
- function: llm-inference, llm-training, chatbot, http-server, api-framework, machine-learning
- domain: large-language-models, artificial-intelligence, chatbots
- platform: python, windows
- tags: vicuna, chatbot-arena, openai-compatible-api, model-serving, llm-evaluation, fine-tuning, distributed-serving, natural-language-processing, linux, macos, docker, gpu, web-server

## Member repositories
- lm-sys/FastChat (main) score 57

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:12:07.893004+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:24:07.316733+00:00, confidence not recorded.
  - readme: https://github.com/lm-sys/FastChat (fetched 2026-08-28T04:12:07.893004+00:00, sha 7968f9409741)
- Data as of 2026-08-30T08:39:29.467469+00:00.