Skip to content
Trending storyDeveloping

Arena ranks Meta's Muse Ember (Max) most cost-efficient in Agent Arena

2 articles1 sourcesince Nov 11Last article

Overview

AISummary of 2 articles

Arena reports that Meta's Muse Ember (Max) leads Agent Arena on cost-efficiency, at about $0.02 per task with a +1.7% net improvement across 6k real-world agentic sessions.

Arena says Ember costs roughly 4x less per task than GPT-6 Luna (Max) and Claude Haiku 5.5, with performance on par across the three.

Arena also reports that Ember finishes about 12x faster than Luna (Max) and 3.5x faster than Haiku 5.5, using 72% and 52% fewer output tokens. A separate Arena post ranks Ember #25 overall, level with Muse Wonder (Max) at #24.

Written by AI from the articles below · updated Nov 11, 3:19 PM ET

Check the sources:

Developments

2 developments

  1. Nov 11, 3:06 PM ET · 1 article
    Arena reports Muse Ember (Max) ranking in Agent Arena
    Arena: Muse Ember (Max) ranks #25 in Agent Arena, level with Muse Wonder (Max)
  2. Nov 11, 3:06 PM ET · 1 article
    Arena ranks Meta's Muse Ember in Agent Arena
    Arena: Arena ranks Meta's Muse Ember as the most efficient model in Agent Arena

Article timeline

The articles in this story. Times are ET.

Nov 11
  1. Arena.aiOfficial
    Muse Ember (Max) ranks #25 in Agent Arena, level with Muse Wonder (Max)

    AIArena reports that Muse Ember (Max) ranks #25 overall in Agent Arena. It performs on par with Muse Wonder (Max), which ranks #24.

    Image from @arena's post
  2. Arena.aiOfficial
    Arena ranks Meta's Muse Ember as the most efficient model in Agent Arena

    AIArena reports that Meta's Muse Ember (Max) leads Agent Arena on cost-efficiency, with a +1.7% net improvement at about $0.02 per task across 6k real-world agentic sessions. Arena says it is roughly 4x cheaper per task than GPT-6 Luna (Max) and Claude Haiku 5.5, with performance on par across the three. Arena also reports that Ember finishes about 12x faster than Luna (Max) and 3.5x faster than Haiku 5.5, using 72% and 52% fewer output tokens.

    Image from @arena's post

Heat Trend

Current comparable heat
6
Peak comparable heat
10Nov 11, 4:00 PM ET
Change in past 24 hours
–

The trend compares only continuously observed participants, so its scope may be smaller than the current heat total. Hover or tap for hourly heat; use arrow keys with a keyboard.