**[Meta's New Model Practically a Giveaway, 29% Cheaper Than DeepSeek]** According to monitoring by Beating, Meta has introduced a low-cost Contributor tier for Muse Spark 1.2 and Muse Code. Users who allow Meta to use their prompts and model outputs to train subsequent models can enjoy significant discounts. The prices for input, cache hits, and output under this tier are $0.10, $0.002, and $0.20 per million tokens, respectively. In comparison, DeepSeek V4 Flash charges $0.14, $0.0028, and $0.28 for the same, making Muse approximately 29% cheaper across all three categories. For users unwilling to share data, the standard tier must be used, where the prices for input, cache hits, and output rise to $1.25, $0.15, and $4.25, respectively—12.5 times, 75 times, and 21.25 times higher than the Contributor tier. The Contributor tier also has lower usage limits, capped at 60 requests per minute and 2.1 million tokens. In contrast, the standard tier supports up to 3,000 requests per minute and 4 million tokens. DeepSeek recently announced a significant price hike for its API, though specific pricing details have yet to be revealed. Meta has already undercut DeepSeek's current prices by 29%, signaling that Zuckerberg is stepping into the role of the price slasher. [Original Link]