A slew of new large language models (LLMs) are set to launch this week, with GPT-6 Sol, Opus 5.5, and Grok 4.7 all waiting in the pipeline.

Beating AI News Flash: This week may bring a flurry of large language model (LLM) updates, with at least 7 models showing signs of release or testing. Tier 1: GPT-6 Sol, Claude Opus 5.5. OpenAI has not confirmed GPT-6 Sol, but Sam Altman previously noted the "major product" originally scheduled for launch last week was delayed to this week; recent community leaks have pointed to GPT-6 Sol testing and gradual rollouts. For Anthropic, blogger Lyra said Opus 5.5 is being tested under the code name claude-wafer-eap, with a planned Tuesday release. Reuters reported last week that Anthropic is considering a new model launch soon, but did not confirm the Opus 5.5 name. Tier 2: Grok 4.7, Qwen4, Kimi K3.1. Elon Musk initially said Grok 4.7 would launch around September 12, later adding it would take "a few more days", and it remains unreleased. Qwen’s developer has clarified that Qwen3.8-Flash-Next is an early preview of the Qwen4 architecture, and the Yunqi Conference is set for September 22-24, though no official announcement plans for Qwen4 at the event have been made. Kimi recently teased K3.1 via a π hint, but no release date, model specifications, or API details are available yet. Tier 3: Gemini 4 Pro, DeepSeek V4.1 Pro. Gemini 4 Pro has been leaked recently to be tested under code names like Argon on Arena, but Google has only confirmed Gemini 4 is in development, with no release date disclosed. DeepSeek has directly referenced the upcoming V4.1 Pro in its official documentation, but also has not provided a launch timeline.