LMArena icon
LMArena icon

LMArena

Public web platform for large language model evaluation using anonymous pairwise comparisons, crowd-sourced voting, real-time model identity reveals and aggregate performance tracking for both open-source and proprietary LLMs across community-submitted prompts.

LMArena screenshot 1

Cost / License

  • Free
  • Proprietary

Platforms

  • Online  LMArena is fully web-based and works on all modern browsers across desktops, laptops, tablets, and mobile devices. No downloads or installations are required to explore and test AI models.
-
No reviews
11likes
0comments
0news articles

Features

Suggest and vote on features
  1.  Dark Mode
  2.  No registration required
  3.  Ad-free
  4.  Text to Image Generation
  5.  Anonymous voting
  6.  AI Writing
  7.  AI Chatbot

 Tags

  • large-language-model-tool
  • artifical-intelligence
  • large-language-model
  • large-language-model-evaluation
  • llm-apps
  • llm-evaluation
  • language-model-evaluation
  • language-models

LMArena News & Activities

Highlights All activities

Recent activities

Show all activities

LMArena information

  • Developed by

    US flagLMArena
  • Licensing

    Proprietary (Apache-2.0) and Free product.
  • Written in

  • Alternatives

    23 alternatives listed
  • Supported Languages

    • English

AlternativeTo Categories

AI Tools & ServicesPhotos & Graphics

GitHub repository

  •  918 Stars
  •  119 Forks
  •  10 Open Issues
  •   Updated  
LMArena was added to AlternativeTo by o_o on and this page was last updated .
No comments or reviews, maybe you want to be first?
Post comment/review

Featured in Lists

User-Friendly, Open Source, Security Focus, Privacy Focus, Cross-Platform

List by almondelf2 with 42 apps, updated

What is LMArena?

LMArena is a public, web-based platform that evaluates large language models using anonymous, crowd-sourced pairwise comparisons. Users submit prompts to two anonymous models, compare their responses, and vote for the one they think performs better. After voting, the models’ identities are revealed, allowing the community to track and analyze global model performance over time.

Official Links