Meta launches Muse Ember for low-cost high-volume AI tasks

A new entry in the AI race

Meta officially released a new artificial intelligence model on Monday called Muse Ember. This launch comes as the company faces increasing competition from other major firms like Anthropic and OpenAI. The model was developed by Meta Superintelligence Labs to serve as a lightweight option for large-scale operations.

The company intends for Muse Ember to be its most cost-efficient model to date. By focusing on affordability, Meta aims to provide a solution that does not sacrifice performance for price. It is positioned as a tool for high-volume work where cost-per-token is a primary concern for developers.

As reported by Times Now, the model was launched alongside other new releases from competitors. Meta is positioning Muse Ember as a way to make large-scale AI operations more viable for businesses. It seeks to balance the need for intelligence with the practical reality of operational budgets.

Versatility across media types

One of the primary technical features of Muse Ember is its ability to process multiple types of data. Beyond standard text, the model can interpret images and audio files. This multi-modal capability means that a single model can handle various types of business inputs simultaneously.

For companies, this versatility simplifies the technical architecture of AI agents. Instead of needing separate models for different formats, an agent can use Muse Ember to analyze screenshots, charts, and scanned documents. It can also process voice notes, meeting recordings, and videos within the same framework.

This unified approach is particularly useful for processing complex documents that contain mixed media. By handling these formats in one place, the model reduces the complexity of building automated systems. It allows for a more streamlined flow of information from various sources into a single processing pipeline.

Targeting repetitive AI tasks

Meta designed Muse Ember specifically for tasks that AI agents repeat all day. These include

Pricing and market competition

The pricing of Muse Ember is a central part of its market strategy. Meta has set the cost at 10 cents per million input tokens. This price point is intended to be competitive against both premium models and open-source alternatives.

Meta used a composite benchmark score from Artificial Analysis to compare its new model against rivals. The company claims that Muse Ember offers advantages over OpenAI’s Luna and Anthropic’s Haiku. It also positions the model as a competitive option against open-source models like DeepSeek V4.1 Flash and GLM 5.3 Flash.

By targeting a lower price point, Meta is attempting to capture the market for high-volume, low-complexity tasks. This allows users to run more frequent queries and process more data for less money. The goal is to make it easier for developers to integrate AI into every step of a digital workflow.

Availability and integration

Users can begin accessing Muse Ember immediately through several different platforms. It is available today on the Meta Model API and within the Muse Code environment. These platforms provide the infrastructure for developers to build and deploy their own applications using the model.

In addition to the standard Meta platforms, the company has expanded the availability of its new models to the cloud. Zuckerberg announced that Muse Ember, Muse Spark 1.3, and Muse Wonder are all available on Google Cloud. This move ensures that a wider range of enterprise customers can access the technology.

The release of these models marks a significant step in Meta's effort to provide diverse AI options. By offering different models for different needs, Meta allows users to choose the best balance of cost and power. Muse Ember serves as the entry point for those who need to process massive amounts of data cheaply.