Anthropic, OpenAI Release Rival Models Minutes Apart as Meta's Muse Surges

Anthropic and OpenAI released competing model updates within roughly 90 minutes of each other, with Anthropic rolling out Opus 5.5 followed shortly by OpenAI's GPT-6 model updates. The back-to-back launches turned the week into an unofficial "model drop week" for the two rivals, an outcome that sits awkwardly against earlier talk from AI leaders at both companies about "pacing the frontier."

The company drawing the most attention, however, was Meta. Its personal AI agent, Muse, is reportedly outpacing ChatGPT's early adoption numbers and is being positioned for smart glasses, suggesting Meta sees the agent as a companion-style product rather than a purely text-based assistant.

The launches extend a period of rapid model turnover across the industry. Earlier releases in this cycle included start-up TypeSafe AI's "Jev" model, aimed at faster and more efficient developer workflows, and Xiaomi's open-weight MiMo V2.6 Pro and Flash models, which showed competitive benchmarks and cost efficiency against DeepSeek V4.1 Flash. Not all updates have landed well: Google's Gemini has drawn user criticism over inaccuracies and difficulty following instructions, including among paying subscribers, while researchers continue testing techniques such as continued pretraining and retrieval-augmented generation to give local large language models domain-specific knowledge.