Recent model releases and agent capability benchmarks have kept trader consensus tightly balanced among OpenAI, Anthropic, and Google DeepMind for second-place status by year-end. OpenAI’s September GPT-6 Astra and always-on Dots agents expanded tool-use reach across thousands of apps, though safety incidents prompted training pauses. Anthropic countered with Claude Opus 5.5 and a new marketplace for connectors, strengthening its position in long-context coding agents. Google advanced Gemini Agent Mode and Skills features while integrating voice and cloud-migration agents. Market-share data shows OpenAI leading overall usage but with Anthropic and Google gaining ground in enterprise and specialized workflows, leaving room for any lab to pull ahead on the next major release or benchmark improvement before December.