E-Ink News Daily

← Back to list

Introducing Our Smallest, Most Efficient Model: Muse Ember

Meta AI has launched Muse Ember, their smallest and most efficient model targeting the same category as Claude Haiku and GPT Luna. Priced competitively at $0.10/million input tokens and $0.50/million output for short prompts, it's positioned on the Pareto efficiency frontier per Artificial Analysis benchmarks.

Background

Meta AI continues to expand its small-model lineup to compete with OpenAI and Anthropic in the cost-efficient, high-throughput segment of the market.

Source
Simon Willison
Published
Nov 10, 2026 at 03:28 PM
Score
6.0 / 10