Salesforce Introduces Cost-Effective AI Model Koa Amid Growing Partnership with Anthropic

Salesforce has introduced its own AI model, Koa, in collaboration with Nvidia. This development comes just three weeks after the company announced a major partnership with Anthropic to make Claude the default model.

Koa is a CRM reasoning model that is based on Nvidia's open Nemotron model and has been trained on a large amount of synthetic data and CRM processes from Salesforce's 27 years of experience. It was introduced as part of the Salesforce Trusted Enterprise AI Harness, which ensures actions, context, security, and AI models come together correctly to execute actions on the platform.

According to Rohan Kumar, President, Chief Platform and Engineering Officer at Salesforce, 'The Agentic Enterprise won't be defined by which model a company chooses. Models will continue to change, and intelligence will increasingly be available everywhere. What will differentiate an enterprise is the trusted, proprietary context it brings to that intelligence.'

Koa is positioned as a more cost-effective alternative to other AI models, particularly for simple conversations and questions. It has been shown to match or exceed leading model performance on CRM actions with 3x fewer errors, although further analysis suggests its performance falls short of some competitors.