aifollow.news
Back Rohan Paul
Rohan Paul· @rohanpaul_ai · X· · Original publication time AI score54

Meta releases small Muse Ember model on its API and Muse CodeMachine translation

AI introduction

Muse Ember targets repeated agent tasks such as tool calls, coding subtasks, routing and extraction. Meta says it matches Claude Haiku 5.5’s intelligence at about a third of the cost, although the material says their rates are identical for prompts under 100K tokens and attributes Ember’s cost advantage mainly to using fewer tokens per task.

Original post · AI translation

@rohanpaul_ai
Translation · refer to the original

Meta 发布了其最小的模型 Muse Ember。Meta 称,该模型的智能水平与 Claude Haiku 5.5 相当,而成本约为后者的三分之一。

面向智能体整天反复执行的工具调用、编码子任务、任务路由和信息提取。

> 不超过 200K 个 token:每 1M 个输入 token 收费 $0.10,每 1M 个输出 token 收费 $0.50 > 超过 200K 个 token:每 1M 个输入 token 收费 $0.30,每 1M 个输出 token 收费 $1.50

对于低于 100K 个 token 的提示词,Haiku 的收费标准相同,因此 Ember 的成本优势主要来自每项任务消耗的 token 更少。

QuoteMark Zuckerberg@finkd
Translation
今天发布 Muse Ember——我们最小的模型,单位成本所提供的智能水平高于同类其他模型。专为智能体整天反复执行、消耗大量 token 的工作而打造:工具调用、编码子任务、任务路由、信息提取和长文档解析。现已在 Meta Model API 和 Muse Code 上推出。
Original

Releasing Muse Ember today -- our smallest model that delivers higher intelligence per dollar than others in its class. Built for token-heavy work that agents repeat all day: tool calls, coding subtasks, routing, extraction, and long document parsing. Available today on the Meta Model API and in Muse Code.

View this post on X ↗

Source: Rohan Paul · x.com

Model
Found an error? Send a correction