Anthropic introduced its Sonnet 4.6 model update as the new default in Claude.ai and Claude Cowork for free and pro users, emphasizing a deeper focus on enterprise Artificial Intelligence. Sonnet 4.6 is described as having improved coding abilities, with better consistency, enhanced computer use skills, and stronger instruction-following. Anthropic stated that the model is not yet as good as humans in computer use, but that its rate of progress continues to accelerate, and it is particularly strong at reading context before modifying code.
The launch, however, landed in the shadow of controversy involving OpenClaw, an open source framework that lets users build personal Artificial Intelligence agents on their own hardware. OpenClaw, originally named Clawdbot, had directed users to Claude Opus, urging them to run projects with Claude 3.5, but Anthropic objected to the Claude-themed naming and pushed developer Peter Steinberger to rebrand first to Moltbot and then OpenClaw. The forced renaming created an opening for scammers to hijack OpenClaw’s account and preceded Steinberger’s hiring by OpenAI, while OpenClaw remains open source as part of a foundation there. Analysts argued that the episode undercuts Anthropic’s positioning as an enterprise-friendly partner and that any technical gains in Sonnet 4.6 are being eclipsed by what they called a public relations setback. One analyst said that Anthropic could have released AGI [artificial general intelligence] this week, and that news would still be overshadowed by the OpenClaw PR disaster.
Despite the backlash, industry analysts see Sonnet 4.6 as signaling Anthropic’s transition from a pure model provider to an agentic solutions and platform company. Claude 4.6, which includes Opus and Sonnet, is described as enabling autonomous execution of multi-step business workflows with good accuracy. By expanding its context window to one million tokens, the updated model is said to let enterprises process entire codebases or legal archives with a single prompt, while adaptive and extended thinking on the Claude Developer platform is framed as a way to balance high-level reasoning with cost efficiency. Integration with model context protocol connectors for tools such as Excel is presented as evidence that Sonnet 4.6 can interact with enterprise data in a controlled way, and analysts conclude that Anthropic is clearly moving toward becoming an enterprise Artificial Intelligence applications and platform vendor, even if its capabilities are still incomplete.
