AI Trends

Arena: Meta Muse Ember Rewrites the Code Arena WebDev Cost-Effectiveness Frontier

ArenaUpdated Nov 17 03:49

AI summary

Arena says Meta's Muse Ember has rewritten the Pareto frontier of Code Arena: WebDev. The model scores 1608 at about 3 cents per answer, whereas the previous best at the same price was 1438; on that basis it would win about 73% of matchups. Arena also lists five models that entered the frontier during the same period, including Claude Fable 5.2 (xHigh), Gemini 4 Carbon (High), Kimi K3.1 Pro (Max), GPT-6.1 Terra (Max) and Muse Ember (Max).

Why read this

SelectedAI score 67

Arena uses a price-versus-score comparison to show where five new models stand on the WebDev leaderboard, which can be used to judge cost-effectiveness differences at the same price point.

Topics

Tags

Quoted post

Arena.ai

@arena

As the top scores move 3x faster, April’s best costs 20x less. https://t.co/oKOvVP2X3A

View original

As the top scores move 3x faster, April’s best costs 20x less. https://t.co/oKOvVP2X3A

View original post

Body

Muse Ember by @AIatMeta has reshaped the Pareto frontier for Code Arena: WebDev.

In our last article, we put the AI race under the spotlight and tracked how price-performance trade-offs improved on Code Arena: WebDev from April to October 2026.

But competition never sleeps. Since then, five new models have joined the frontier: - Claude Fable 5.2 (xHigh) by @AnthropicAI - Gemini 4 Carbon (High) by @GoogleDeepMind - Kimi K3.1 Pro (Max) by @Kimi_Moonshot - GPT-6.1 Terra (Max) by @OpenAI - Muse Ember (Max) by @AIatMeta

Of the five, Muse Ember moved the frontier the most. At about 3 cents an answer, it scores 1608, against 1438 for the previous best at that price. That means it would win about 73% of the time.

Curious how we got here? Full article below.

Source

Arena