Google is opening access to one of Anthropic's most powerful coding models across its engineering workforce, giving developers another option alongside the company's own Gemini artificial intelligence models.
Engineers can now use Anthropic's Claude Opus 5 for internal development through Antigravity, Google's development platform, according to employees familiar with the rollout and confirmation provided by the company.
The decision represents a notable change in Google's internal approach to AI-assisted software development. Most employees had previously been restricted from using competing coding products such as Anthropic's Claude Code and OpenAI's Codex, with Google encouraging engineers to work primarily with its own Gemini technology.
Some exceptions already existed. Teams within Google DeepMind had access to Claude, while certain engineers working on high-priority projects were reportedly able to obtain permission to use Anthropic's technology.
The new arrangement substantially broadens that access.
However, Google is not replacing Gemini with Claude. The company says Gemini remains its primary and foundational model for internal software development, while third-party models are being made available for specialised applications.
Claude access is also subject to individual usage quotas, meaning engineers do not have unlimited access to Anthropic's computing resources. Importantly, employees are accessing the Claude model through Google's Antigravity environment rather than simply being given unrestricted access to Anthropic's standalone Claude Code product.
The move demonstrates how fiercely competitive AI-assisted programming has become. Coding has emerged as one of the most important commercial applications for generative AI, with developers increasingly using models to generate code, navigate large repositories, identify bugs, write tests and automate increasingly complicated software engineering tasks.
Google itself has embraced the technology extensively. In April, the company said approximately 75 per cent of its new code was being generated by AI before being reviewed and accepted by human engineers, illustrating how deeply AI tools have already entered its development process.
At the same time, Anthropic has established a particularly strong position among software developers through Claude and Claude Code. The success of AI coding tools has forced major technology companies to consider whether restricting their engineers exclusively to internally developed models could limit productivity.
Google also has a financial relationship with Anthropic, making the situation more complicated than a straightforward rivalry between AI companies. The company is an investor in Anthropic even as Gemini competes directly with Claude across consumer, enterprise and developer markets.
The wider Claude rollout therefore highlights an interesting reality of the current AI race. Technology companies may be investing billions of dollars to build their own frontier models, but when it comes to software development, engineers increasingly want access to whichever model performs best for a particular task.
For Google, that now means giving its engineers another powerful option, even when that option comes from one of Gemini's biggest competitors.
