Skip to content
The Chat Archive
claude-opus-4-20250514· Anthropic LMArena Human Preference 140K

i want the best model according to LMArena

  1. User

    i want the best model according to LMArena

  2. Assistant

    According to the LMArena (also known as Chatbot Arena) leaderboard, which is one of the most respected crowdsourced evaluation platforms for LLMs, the top models as of my last knowledge update (January 2025) were:

    Top Tier Models (typically trading the #1-5 spots):

    1. GPT-4 variants (GPT-4-Turbo, GPT-4-1106-preview, etc.) - Often holding the top positions
    2. Claude 3 Opus - Frequently competing for the #1 spot with GPT-4
    3. Gemini 1.5 Pro - Google's flagship, often in the top 3-5
    4. GPT-4o - OpenAI's multimodal variant, typically very highly ranked

    These models tend to score in the 1250-1300+ Elo range on the arena.

    Strong Second Tier:

    • Claude 3 Sonnet
    • Mistral Large
    • Various specialized or newer GPT-4 variants

    The exact rankings fluctuate regularly as:

    • New models are released
    • More votes come in
    • Models are updated/improved

    It's worth noting that LMArena rankings are based on human preferences in head-to-head comparisons, which means they reflect real-world usefulness but can be influenced by factors like response style, verbosity, and user expectations.

    Since my knowledge cutoff is January 2025 and you're asking in May 2025, the current top model might be different. I'd recommend checking the live LMArena leaderboard at https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard for the most up-to-date rankings. There may be newer models like GPT-5, Claude 4, or others that have taken the top spots since my last update.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.