December 2025 Summaries

DigitalOcean has introduced native support for the Bun framework on its App Platform, allowing developers to deploy Bun applications directly from their code repositories without requiring configuration. Bun, a high-performance JavaScript runtime, is noted for its faster startup times and lower memory usage compared to Node.js, making it a desirable option for modern applications. The integration with App Platform utilizes Cloud Native Buildpacks to automatically detect, build, and run Bun applications, providing seamless support for frameworks like Next.js. Developers can choose from multiple deployment paths, including using Dockerfiles or pre-built images, and can easily migrate existing Node.js applications to Bun by updating project metadata. The platform also offers caching for faster builds and supports specifying Node.js versions when Bun is used as the package manager. Overall, the addition of Bun support enhances the performance and ease of deployment for JavaScript applications on DigitalOcean's App Platform.

In 2025, DigitalOcean Managed Databases introduced several enhancements aimed at increasing power, flexibility, and simplicity for developers and businesses. Key developments included performance boosts, new database engines, automation, and observability improvements, all geared toward facilitating the creation and management of reliable data-backed applications. Notable releases included support for new versions of PostgreSQL and MongoDB, expanded storage capabilities, the introduction of the DigitalOcean Managed Caching for Valkey, and the implementation of Role-Based Access Control. The latter half of the year saw the launch of the MCP Server, which allows cloud resource management through AI-powered tools, advanced configuration options in the Managed Databases UI, and storage autoscaling to prevent downtime. Additionally, DigitalOcean supported Remote MCP Server connections for AI tools, improved migration tooling, and enhanced reliability with features like automatic restarts for MongoDB clusters and MySQL incremental backups. These advancements reflect DigitalOcean's commitment to optimizing database management and infrastructure workflows, while providing seamless, secure, and efficient solutions for their users.

DigitalOcean has enhanced its Gradient™ AI Platform with new features for building production-ready retrieval-augmented generation (RAG) systems, now available in public preview. The platform's code-first approach allows developers to create, manage, and query knowledge bases entirely through code, offering complete control over data ingestion, chunking, embedding, and retrieval without the need to manage infrastructure. The improvements include direct API access for seamless integration, customizable content ingestion from various sources, flexible chunking and embedding strategies, and advanced retrieval techniques with citation-backed answers. These features aim to address the challenges developers face in scaling and customizing knowledge bases for production workflows, providing a comprehensive toolkit that supports natural language queries, metadata filters, and hybrid search capabilities. The enhancements are designed to streamline the development of smarter AI applications by enabling developers to transform data into actionable, context-rich answers efficiently.

Developers often struggle to transition AI agent prototypes into reliable, production-ready systems due to the complexity of orchestrating interactions, managing state, and deploying tools. The DigitalOcean Gradient™ AI Agent Development Kit (ADK) aims to streamline this process by providing a robust, code-first SDK that enables the creation, testing, and deployment of sophisticated agent workflows directly within existing development environments. This toolkit offers features such as multi-step workflow orchestration, state management, tool integration, and knowledge base support, while also facilitating evaluations, tracing, and deployment through a consistent workflow. The public preview introduces enhanced capabilities, including detailed tracing for understanding agent behavior, knowledge base integration for contextual accuracy, and evaluations for multi-step agents, along with upcoming features like Agent-to-Agent communication. The ADK is designed to simplify the setup process, allowing developers to efficiently build and deploy AI agents by consolidating various functions into a single command.

DigitalOcean has introduced a new custom date range billing view feature, available at no additional cost to its customers, to enhance transparency and management of cloud costs. Accessible via the Billing Console under Billing → Insights, this feature allows users to filter billing data by custom date ranges, providing daily, weekly, and monthly breakdowns of spending across different products such as Droplets, Databases, Spaces, and Bandwidth. This granularity aids DevOps and finance teams in proactively managing and aligning budgets with project timelines, fiscal quarters, and internal departmental costs, while also enabling quick detection of cost anomalies. The feature supports organization-level reporting by including financial adjustments like credits and taxes, offering a comprehensive view of expenses. Tailored for growing businesses with expanding infrastructure, this update marks the first step towards more advanced cost analytics and optimization tools, reflecting DigitalOcean's commitment to improving cost insights and transparency for technology companies.

DigitalOcean is set to enhance its GPU offerings by introducing GPU Droplets powered by NVIDIA HGX™ B300, which represents a significant advancement in AI computing capabilities. The NVIDIA HGX B300, part of the Blackwell Ultra accelerated computing platform, is designed to handle complex AI workloads with improved computational power, memory bandwidth, and energy efficiency. This innovation supports demanding applications such as generative AI, data analytics, and high-performance computing by providing 1.5 times more dense Tensor Core FLOPS and expanded memory. The integration of NVIDIA Spectrum-X Ethernet networking into DigitalOcean's ecosystem will also optimize AI throughput, offering up to 1.6 times higher performance than standard Ethernet, thus reducing training times for large models. Additionally, DigitalOcean's approach of providing pre-configured instances with essential AI/ML frameworks and its commitment to cost-effectiveness, seamless integration, reliability, and observability make this a compelling option for developers aiming to push the boundaries of AI development.

In 2025, DigitalOcean made significant advancements in its Managed Kubernetes service, focusing on simplifying operations, enhancing security, and increasing scalability. Key updates included engine upgrades, networking and security enhancements, and ecosystem integrations, which collectively reduced operational overhead for users. Major releases throughout the year introduced features like increased cluster capacity, VPC-native networking, eBPF-powered routing, Managed Cilium for security and observability, and new GPU Droplet types for AI and machine learning workloads. The year also saw the introduction of a new AI-optimized data center in Atlanta, support for the DigitalOcean MCP Server for natural-language cloud management, a managed Gateway API for advanced traffic management, and features like VPC NAT Gateway and Network File Storage for secure and efficient operations. Additionally, DigitalOcean expanded support for multi-node GPU configurations, facilitating scalable, high-performance applications, and announced a variety of upcoming initiatives for 2026, including webinars and a migration program.

DigitalOcean has introduced remote support for its Model Context Protocol (MCP) Server, allowing developers to connect AI tools to DigitalOcean services without the need for local installations. The remote MCP endpoints are available for nine DigitalOcean services, such as App Platform, Databases, and Kubernetes, each running as a standalone MCP server accessible via dedicated HTTPS URLs. This advancement simplifies the setup process, requiring only updates to the MCP client configuration to point to the hosted endpoints and the inclusion of a DigitalOcean API token for authentication. The remote MCP offers benefits including no local dependencies, modular connections, and automatic updates, which facilitate standardized configurations across teams. Existing tutorials and documentation remain applicable, with changes limited to updating the MCP client to reference remote endpoints. This evolution aims to streamline AI-powered workflows by reducing setup complexity and enhancing manageability of DigitalOcean resources.

Jeff Fan, a Senior Solutions Architect at DigitalOcean, leverages his technical acumen and customer-focused approach to assist businesses in Europe and Asia in solving complex cloud issues using DigitalOcean's solutions. Initially attracted to DigitalOcean for its user-friendly platform, Fan finds his role as a trusted technical advisor to be essential in bridging customers' current needs with their future aspirations. The company's rapid innovation pace, driven by customer feedback, and its commitment to customer-centricity, particularly in fast-moving APAC markets, underscore its momentum and growth. Fan's personal experiences with the supportive culture at DigitalOcean, including during a family bereavement, highlight the company's emphasis on empathy and community, reinforcing its philosophy of simplicity and ease of use. DigitalOcean's environment is portrayed as one that values continuous learning, bold thinking, and strong interpersonal connections, making it an attractive workplace for those passionate about cloud technology.

DigitalOcean has introduced DoTs, a new TypeScript SDK designed to enhance developer interaction with its resources by utilizing TypeScript's type safety and modern features. The development of DoTs leverages automated code generation using OpenAPI, GitHub Actions, and Kiota, significantly reducing manual effort and errors in SDK creation. This automation ensures that the SDK remains up-to-date with API changes and supports robust testing and documentation processes. Kiota, an open-source API client code generator, aids in producing strongly typed SDKs across various languages, emphasizing consistency and ease of maintenance while offering flexible authentication options. The integration of CI/CD workflows and comprehensive testing, including mocked and integration tests using Jest, ensures the reliability of the SDK, while automated documentation is maintained through TypeDoc and Read the Docs. This initiative aims to streamline development workflows, promoting reliability, scalability, and user-friendly experiences for developers managing DigitalOcean resources.

The DigitalOcean Gradient™ AI Platform has introduced updates to its agent evaluations feature, aimed at enhancing the speed and effectiveness of AI agent assessments. The redesigned evaluation experience addresses previous challenges by introducing goal-oriented metric grouping, example datasets, and clear, persistent error messaging, which simplifies the debugging process. Metrics are organized into intuitive groups like Safety & Security and Correctness, with the former preselected for quick startup. Deep integration with observability tools allows developers to trace low scores back to the source for precise debugging and improvements. These evaluations help developers systematically test and optimize AI agents, providing insights into performance and enabling faster, more reliable deployment. The platform offers a step-by-step tutorial for new users to create test cases, select metrics, and interpret results, facilitating the development of safer and more efficient AI systems.