Meta Superintelligence Lab released Muse Ember on Monday, positioning it as the company’s lightest and most cost-efficient AI model to date. The release states Ember is designed for complex reasoning execution and repetitive agent work such as tool calls, coding subtasks, routing, extraction, and long document parsing. Meta highlighted the model’s cost-per-task using Artificial Analysis’s composite benchmark score to show it is cheaper and smarter than competitors like Anthropic’s Haiku and OpenAI’s Luna. At 10 cents per million input tokens, Ember is significantly cheaper than frontier competitors while remaining cost-competitive with open-source options like DeepSeek V4.1 Flash and GLM 5.3 Flash.
Given the emphasis on low cost and accuracy at scale, Ember is designed to pair with a larger model like Meta’s Muse Wonder in an agentic setup. In a customer-support scenario, a heavier model might establish rules for handling an issue before sending a lighter-weight model to apply those rules across tens of thousands of cases. Meta notes that Muse Wonder plans while Muse Ember handles the volume. The model also trained to work well across text, images, and audio, allowing agents to read screenshots, charts, and scanned documents or work from voice memos and meeting recordings without a separate model. Ember is the fourth model under the Muse brand, following Spark, Glimmer, and Wonder, which precedes the flagship consumer product Muse Agent.
- Cost is 10 cents per million input tokens
- Designed to pair with Muse Wonder for volume tasks
- Handles text, images, and audio inputs natively
