Meta is entering the high-performance AI model market with the release of Muse Ember, designed as a direct competitor to models from OpenAI (GPT-6 Luna) and Anthropic (Claude Haiku 5.5).
Meta is entering the high-performance AI model market with the release of Muse Ember, designed as a direct competitor to models from OpenAI (GPT-6 Luna) and Anthropic (Claude Haiku 5.5).
The release was preceded by a long struggle in the affordable technology segment. A few months ago, Chinese laboratories released a line of models that provided about 80-90% of the performance of leading solutions at a significantly lower price. OpenAI responded with the launch of GPT-6 Sol and GPT-6 Luna , and Anthropic last month introduced Haiku 5.5 for processing large amounts of data. These moves have restrained the growth of popularity of Chinese alternatives in the United States.
Now Meta has introduced its smallest and most efficient model, the Muse Ember, which is as intelligent as its competitors, but costs much less, writes Neowin.
In the Artificial Analysis Intelligence Index v5 test, the Ember model scored 36 points, tying with Claude Haiku 5.5 and surpassing GPT-6 Luna (28 points).
At maximum reasoning level, one query in Muse Ember costs about $0.14, while in Claude Haiku 5.5 it costs $0.41, and in GPT-6 Luna it costs $0.15. This gives Haiku-level quality for almost three times the price.
Claude Haiku 5.5 still holds the lead in the FrontierCode 2.0 benchmark (an assessment of real-world software engineering tasks), but Ember supports multimodality — text, image, audio, and video input. The context window is 1 million tokens, and the maximum generation volume is up to 512,000 tokens.
The price for Muse Ember is $0.10 per million input tokens, $0.004 per million cached tokens, and $0.50 per million output tokens (for prompts up to 200,000 tokens). For longer queries, the price increases to $0.30 per input token and $1.50 per output token.
The model is already available via the Meta Model API and Muse Code (ID: muse-ember-1.0). It supports the OpenAI Responses API specification, allowing developers to easily integrate it into existing agent scenarios.
