Skip to content
TheChatArchiveThe Chat Archive
DeepSeekdeepseek-r1-0528· DeepSeekHugging Face LMArena Human Preference 140K

when grok 3 released it was at the top of lmarena for a while. what was the landscape of…

From a Chatbot Arena battleThe same person sent this prompt to two anonymous models at once. Compare it with Googlegemma-3-27b-it’s answer.
  1. User

    when grok 3 released it was at the top of lmarena for a while. what was the landscape of other ais at that time, and how long until something replaced it at the top and what was it?

  2. DeepSeek Assistant

    Grok-3 (specifically Grok-1, released in November 2023) did indeed briefly top the LMSys Chatbot Arena leaderboard around December 2023 - January 2024. Here's the landscape and what happened next:

    Key Competitors at Grok-1's Peak (Late 2023/Early 2024)

    1. OpenAI's GPT-4 Turbo (Nov 2023):
      The dominant leader before Grok-1's rise. Slightly outperformed in coding but trailed Grok-1 in overall Arena Elo.
    2. Anthropic's Claude 2.1 (Nov 2023):
      Competitive in reasoning and long-context tasks, but ranked below Grok-1/GPT-4 Turbo.
    3. Google's Gemini Pro (Dec 2023):
      Initial release underperformed; ranked below top models.
    4. Open-Source Models (Mixtral, Yi, Llama 2):
      Mixtral-8x7B was the strongest open-source model but still trailed top proprietary models by ~100 Elo points.

    How Long Did Grok-1 Stay at #1?

    • Approximately 6–8 weeks (mid-Dec 2023 to early Feb 2024).
    • Grok-1's lead was narrow. By late January 2024, GPT-4 Turbo and Claude 2.1 were statistically tied with it.

    What Dethroned Grok-1?

    • Claude 3 Opus (released March 4, 2024) decisively took #1:
      • Surpassed Grok-1 by ~50 Elo points immediately.
      • Dominated in reasoning, math, coding, and long-context tasks.
      • Remained #1 for months until GPT-4o arrived.

    Timeline Summary

    DateEventLeaderboard Impact
    Nov 2023Grok-1 (Grok-3) releasedReached #1 by mid-Dec 2023
    Dec 2023–Jan 2024GPT-4 Turbo/Claude 2.1 updatesGrok-1 tied with rivals
    Mar 4, 2024Claude 3 Opus released#1 spot taken, Grok-1 fell to #3
    May 2024GPT-4o & Gemini 1.5 Pro releasedGrok-1 dropped further (top 10)

    Why Grok-1's Lead Was Short

    • Rapid Model Releases: Claude 3 and GPT-4o leapfrogged capabilities within months.
    • Narrow Margins: Grok-1's lead was small (<20 Elo over GPT-4 Turbo).
    • Specialized Strengths: Grok-1 excelled in "fun" interactions (aligned with X's platform) but lagged in technical tasks vs. Claude 3/OpenAI.

    As of mid-2024, Claude 3 Opus, GPT-4o, and Gemini 1.5 Pro dominate the top, while Grok-1 fell to #8–#10 (behind models like Claude 3 Sonnet and Mixtral 8x22B). Grok's team is developing Grok-1.5 with improved reasoning, but it hasn't challenged the top tier yet.

    The AI race remains volatile—expect reshuffling with Grok-2, Claude 4, or Gemini 2.0! 🚀

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