Meta's Chief AI Officer, Alexandr Wang, announced that the newly launched Muse Spark 1.3 maintains the same price point as its predecessor, Muse Spark 1.2, while offering enhanced efficiency. Specifically, Muse Spark 1.3 requires 25% fewer tokens for operation. At present, Meta has not finalized its decision regarding the public release of the weights for Muse Spark 1.3. However, the company intends to proceed with making the weights of Muse Spark 1.2 available to the public. Furthermore, Meta is on track with the development of a larger-scale model, codenamed 'Watermelon', and is confident that it will demonstrate strong competitiveness in the market.
