Tabnine is pleased to announce support for Claude 3.7 Sonnet, the latest AI model from Anthropic, in our AI software development platform. This integration reinforces our commitment to delivering high-performance, flexible, and secure AI tools that empower enterprise engineering teams. Claude 3.7 enhances software development workflows with improved coding accuracy and more effective tool interactions, allowing developers to streamline their processes while maintaining full control over their AI stack.
As AI adoption accelerates, many providers are attempting to lock engineering teams into closed ecosystems, restricting their choice of models, development environments, and infrastructure. Tabnine has always taken a different approach. We believe that true innovation requires flexibility and architectural freedom, ensuring that companies can deploy AI in a way that aligns with their strategic and security priorities.
Since pioneering the AI code assistant category in 2018, Tabnine has led the market in developing industry-first innovations. A year ago, we introduced switchable models, becoming the first AI provider to offer this level of flexibility. Since then, model choice has become a standard expectation across the industry. However, while others now claim to provide model flexibility, many still require deployment on their cloud infrastructure, restricting true control over AI usage. Tabnine ensures that engineering teams can integrate Claude 3.7 alongside proprietary and other third-party models, allowing organizations to maintain full control over their software development process while avoiding vendor lock-in.
Claude 3.7 Sonnet is available to all Tabnine SaaS users starting February 26, 2025, with no additional setup required. Enterprise customers will gain access through a phased rollout. For those using EMT and self-hosted deployments, Claude 3.7 can be enabled through a patch in version 5.15 next week, with full availability in version 5.16 on March 18. Enterprise users can integrate Claude 3.7 through either a private endpoint or the Tabnine endpoint via the admin console, ensuring secure and seamless access that aligns with organizational compliance and infrastructure requirements.
With this integration, Tabnine remains at the forefront of AI-powered software development, becoming one of the first platforms to offer Claude 3.7 to enterprise engineering teams.
Claude 3.7 delivers a significant performance boost, making it particularly effective for organizations that manage large and complex codebases. The model has achieved 70.3 percent accuracy on the SWE-bench Verified benchmark with custom scaffolding, surpassing previous models in real-world software engineering tasks. This improvement enables developers to tackle complex coding challenges, optimize large-scale refactoring efforts, and automate debugging with greater precision.
The model also excels in agentic tool use, achieving 81.2 percent accuracy in executing automated workflows. This advancement enables organizations to leverage AI for code reviews, test generation, and compliance validation, reducing manual effort and accelerating software development cycles.
Tabnine’s integration of Claude 3.7 is about more than just performance gains. It is designed to meet the rigorous security, privacy, and compliance requirements of enterprise software development. Unlike other AI coding assistants, Tabnine ensures that organizations retain full control over their AI stack and intellectual property without the risk of data exposure.
By incorporating Claude 3.7 within Tabnine’s AI software development platform, enterprises gain access to an advanced context engine that delivers precise, project-specific code recommendations. This system integrates Jira issues, repository data, and workspace context, ensuring that AI-generated code aligns with an organization’s development standards and security protocols. Additionally, code provenance & attribution provides built-in auditability for licensing compliance, which allows teams to safely use any performant third party model in their workflows.
Tabnine is also committed to protecting intellectual property and ensuring AI privacy. Unlike other AI platforms that train models using users’ proprietary code, Tabnine guarantees that no user-generated data is used for model training. This commitment to enterprise-grade security allows organizations to deploy Claude 3.7 in private, self-hosted environments, ensuring that sensitive code and business-critical intellectual property remain secure.
As AI technology continues to evolve, Tabnine remains focused on delivering enterprise-ready AI solutions in a platform that eliminates the need to trade off between performance, privacy, and protection. The integration of Claude 3.7 represents an important step toward helping engineering teams automate workflows, enhance software quality, and accelerate innovation—all while ensuring that organizations retain complete control over their IP and data.
Claude 3.7 is now available for SaaS users with no additional setup required. Enterprise customers can enable access via the admin console after upgrading to version 5.15 next week or 5.16 on March 18. For additional information or support, contact your Customer Success Manager or reach out to Tabnine Support.
As the AI landscape advances, enterprises need AI partners that help them optimize automation, accuracy, and security while maintaining the highest standards of privacy and control. Tabnine is committed to leading this transformation, ensuring that organizations can adopt AI on their terms, without compromise.
If you’re exploring AI coding solutions for your team, contact us today to learn how Tabnine can address your concerns, help you overcome your challenges, and deliver the outcomes you’re looking for-with an AI software development platform uniquely tailored to you.
Explore recent posts
//Explore recent posts
//
18 -min read
3 -min read
6 -min read
More Announcements
//3 -min read
4 -min read
3 -min read