Anthropic PBC today introduced a new Claude Enterprise plan aimed at business customers that will allow them to work with trusted internal knowledge and the company’s artificial intelligence models securely at large scale.
Key features of the new Enterprise plan open up the company’s Claude large language model family with a gigantic 500,000-token context window, more than double the 200,000-token context window currently available. This is the equivalent of hundreds of sales transcripts, dozens of 100-page documents or an entire code repository.
“Our goal here is to enable Claude to empower every team to connect with your company’s knowledge so that you can be the most productive and creative version of yourself at work,” Anthropic product manager Nicholas Lin told SiliconANGLE in an interview. “We’re focused on solving the most common, most practical and the most critical business use cases that directly impact your day-to-day work.”
To assist teams in their work, Claude Enterprise brings multiple key features for team collaboration to the forefront. First is Projects, a tool that allows users to work together alongside the AI and their company’s internal knowledge on projects at the same time while sharing with the rest of the team. This enables the entire team to build with Claude, share the LLM’s outputs and their work and allow the entire team to build off each other’s interactions.
Those interactions with Claude can also be turned into something called Artifacts, which are standalone work products that could include code snippets, flowcharts, SVG graphics, websites or interactive dashboards. Artifacts bring collaborative work alongside Claude to a different level by allowing users to see their creations form and iterate in real time. When they’re done, they can also publish them so that they can be used and remixed within other projects.
For developers, Anthropic is introducing native GitHub integration in public beta mode. Teams will be able to sync GitHub repositories with Claude and brainstorm alongside entire codebases to iterate on new features, coming up with new documentation, refactoring use cases, onboarding new engineers and more.
“I think this especially shines paired with copilot and other types of AI-assisted software programs,” Lin said. “This enables you to understand the overall landscape for software development, especially with all the open-source libraries you do not have to copy and replicate into your local development flow.”
The Enterprise plan also introduces critical administrative and security controls for users designed to prioritize the management and protection of critical data. With single sign-on and domain capture, users can securely manage and provision user access and the inclusion of role-based access and fine-grained permission management can enhance data management. Audit logs will also allow users to trace system activities for security and compliance monitoring, which will become available in the coming weeks.
Early customers of the Enterprise plan include GitLab Inc., management consulting firm North Highland and Boston Consulting Group Inc. They have used Claude across a variety of stages of work to collaborate with team members to create outputs and boost productivity and streamline internal processes.
“Since we’ve been piloting Claude, it has revolutionized our workflows and become our most requested tool,” Luka Anic, senior director, technical AI program and product manager at North Highland. “Claude has dramatically accelerated everything from content creation to data analysis. Within a matter of months, our team has unlocked thousands of hours to focus on high-impact initiatives that were previously out of reach.”
Of course, Lin pointed out that Anthropic employees – even himself – make use of Claude daily in their everyday work to lighten their tedium and workloads.
“I use Claude every single day because it’s a fantastic brainstorming partner for myself,” Lin said. “Because it is hooked up to all of my systems, it’s able to have contextual awareness and contextual knowledge in terms of all of my customer insights, and all of the interactions with my coworkers. And based on those insights, it can help me understand the insights from them.”