88Z AI, 11 November — Meta has released Muse Ember, its smallest model to date, focusing on complex reasoning and agent tasks with support for one million token context and native multimodality. The cost per single task is approximately $0.14, about one-third of Claude Haiku 5.5. While it outperforms the latter in several agent benchmarks, it still lags in real-world engineering coding.
At a glance
- Muse Ember supports a one-million-token context and native multimodality, with a single-task cost of around $0.14.
- The model outperforms Claude Haiku 5.5 in several agent benchmarks but still lags in engineering coding capabilities.
The update
Meta has officially launched Muse Ember, an artificial intelligence model that stands as the company's smallest offering to date. This model is optimized for complex logical reasoning and agent tasks, featuring the capacity to handle a context window of up to one million tokens and providing native support for multimodal inputs. In terms of cost efficiency, Muse Ember performs notably well, with the expense for completing a single task amounting to approximately $0.14. This price point is roughly one-third of that associated with Claude Haiku 5.5. In various benchmark tests focused on agent capabilities, Muse Ember has surpassed the performance of Claude Haiku 5.5. However, in practical engineering coding scenarios, the model's performance still shows a gap compared to competitors, not yet achieving the same level of leading advantage.
Sources & editorial note
This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
- 智东西Source published
Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.