AI Image
Meta releases Muse Ember, a new lightweight model for agent execution
TechCrunch • • business
Key Points:
- Meta Superintelligence Lab launched Muse Ember, its lightest and most cost-efficient AI model, designed for complex reasoning and repetitive tasks such as tool calls, coding subtasks, routing, extraction, and long document parsing.
- Ember is priced at 10 cents per million input tokens, making it significantly cheaper than competitors like Anthropic’s Haiku and OpenAI’s Luna, while maintaining competitive performance with open-source models like DeepSeek V4.1 Flash and GLM 5.3 Flash.
- The model is intended to work in tandem with larger models like Meta’s Muse Wonder, where the heavier model plans and the lighter Ember handles high-volume execution, exemplified in customer-support scenarios.
- Ember is trained to process multiple types of media inputs, including text, images, and audio, enabling it to handle screenshots, charts, scanned documents, voice memos, meeting recordings, and video without requiring separate models.
