LMArena: Meta's Muse Ember is 13x faster and ~3x cheaper than GPT-6 Luna and Claude Haiku 5.5
arena · x · 2026-11-18
LMArena's efficient-model leaderboard has a new leader: Meta's Muse Ember (Max) beats OpenAI's GPT-6 Luna (Max) and Anthropic's Claude Haiku 5.5 (High) on every dimension measured.
- Speed: 27s per task vs 5.8 min (Luna) and 1.9 min (Haiku) — roughly 13x faster
- Performance: +2.3% Net Improvement on the Agent Arena score vs +1.4% and +0.4%
- Cost: 1.9¢ per task vs 7.0¢ and 6.9¢ — less than a third of either rival
The post also includes analysis of Ember's steerability score across Work categories and examples of real-world tasks users are shipping with Muse Ember in the Agent Arena.
More from Models
- OpenAI retakes MathArena top spot as GPT-6.2 Astra hits 79.25% on ArXivLean — sandersted · 2026-11-18
- Every Team to Live-Test Astra 6.2 and Sol 6.2, Inviting Developer Scrutiny On-Stream — danshipper · 2026-11-18
- Kling Image 4.0 ranks #7 in text-to-image and #6 in editing, beating Nano Banana — anthara_ai · 2026-11-18
- RumbleBar: open-source macOS menu bar dictation running Rumble-2-Flash locally at ~40x real-time — ricklamers · 2026-11-18
- Dev lets GPT-6.2 Astra improve a shader from his in-development game — OpenAIDevs · 2026-11-18
- Grok Heavy subscriber downgrades to Premium: third-party models in Cursor billed separately, refunds ignored — alexcovo_eth · 2026-11-18