Embedl sets a new standard for on-device LLM inference, releasing the world’s fastest language models for the edge.

Gothenburg, Sweden — December 8, 2025 Embedl, a Swedish deep-tech pioneer in AI model optimization, announced today FlashHead, an optimization