Salesforce plans to spend $300 million [1] on Anthropic tokens and AI services during the 2026 fiscal year [2].

This investment signals a significant shift in how the CRM giant intends to scale its artificial intelligence infrastructure. By committing a large sum to a specific model provider, Salesforce is prioritizing the rapid integration of advanced generative capabilities over a more fragmented approach to AI adoption.

Marc Benioff, CEO of Salesforce, said the company will use the funds to bring Anthropic’s tools into its product development operations. The strategy focuses on the implementation of coding agents, and other generative AI capabilities to enhance the software's functionality [2].

Benioff said the tools improve the efficiency of the company's internal and external offerings. The spend is targeted toward the 2026 fiscal year [2], reflecting a concentrated effort to remain competitive in the AI-driven enterprise software market.

Salesforce has previously maintained a multi-model strategy, but the $300 million [1] allocation suggests a deepening reliance on Anthropic's specific architecture. The integration of these tokens will allow Salesforce to scale its AI features across its suite of products, including Slack and its core CRM platform, more aggressively.

This move comes as the industry shifts from basic chatbots to autonomous agents capable of writing and executing code. Benioff said the company is going all-in on the technology to drive the next generation of product development [2].

“Salesforce plans to spend $300 million on Anthropic tokens and AI services.”

This expenditure demonstrates a transition from the experimental phase of generative AI to a high-capital deployment phase. By investing heavily in Anthropic's tokens, Salesforce is betting that specialized coding agents will provide a tangible productivity gain and a competitive edge in the enterprise sector, potentially reducing the time required to ship new software features.