Overview
- Musk wrote on July 9 that he was “clearly wrong about Anthropic” and praised its Mythos and Fable models as the best available while predicting a follow‑up Mythos 2.
- Anthropic is a major customer of SpaceXAI’s Colossus 1 under a compute lease that reportedly covers more than 220,000 NVIDIA GPUs and about 300 megawatts, with payments near $1.25 billion a month through May 2029 and either side able to exit on 90 days’ notice.
- Musk publicly rejected the idea that he would use that contract to cripple Anthropic, saying he “would never cut them off in a way that hurt them badly, even as a competitor.”
- Anthropic’s Mythos and Fable faced a national‑security export review after their June release but U.S. Commerce Department restrictions have been lifted, and the models are being discussed for advanced cybersecurity and autonomous‑reasoning uses.
- Anthropic’s confidential U.S. IPO filing reported in June after a large funding round valuing the company near $965 billion raises stakes for markets, infrastructure providers and rivals as SpaceXAI, OpenAI and others push competing models such as Grok 4.5.