Salesforce and Anthropic launch Claudeforce

Salesforce and Anthropic announced an expanded partnership, branded Claudeforce, that embeds each company’s product inside the other’s. A new plugin, Salesforce in Claude, gives Claude 37 prebuilt sales skills — including checking a deal’s health, preparing for a meeting, and updating a pipeline — usable from within Claude rather than by switching to the Salesforce application. The plugin launched with a limited set of pilot customers, with an open beta reported as planned for September.

In the other direction, Claude became a reasoning-model option inside Salesforce’s Agentforce platform and the default model powering Slackbot, alongside a new feature called Claude Tag for surfacing team decisions inside Slack conversations. Salesforce had already taken a stake in Anthropic reported at around $5 billion, and coverage of the announcement said Salesforce expected to spend roughly $300 million on Anthropic’s models this year.

The deal extended a pattern of large enterprise software vendors wiring a frontier lab’s models directly into existing workflow tools rather than building comparable reasoning models themselves, betting that owning the point of use — the CRM record, the Slack channel — mattered as much as owning the underlying model. Additional prebuilt skills for the Salesforce-in-Claude plugin were reported to be planned for later in 2026.