A chronological record of Vultr platform updates, feature releases, and improvements to help users stay informed about changes to the cloud ecosystem.
Vultr release notes provide a transparent view into the ongoing improvements, feature releases, and platform enhancements across our cloud ecosystem. These notes are designed to keep customers, developers, and partners informed about both major releases and incremental updates to Vultr products and services. You’ll find details about new features, performance improvements, bug fixes, API updates, and infrastructure changes.
Introducing Clusters: Clusters are now available on Vultr, providing a streamlined way to deploy and manage groups of compute resources as a single unit. With automated fabric networking, clusters benefit from simplified connectivity, improved scalability, and more efficient management of distributed workloads across multiple instances. See the documentation.
Milan Region Now Available: A new Milan location is now available on Vultr, expanding our global footprint in Europe. This new data center provides improved latency, regional redundancy, and additional capacity for deploying workloads closer to users in Southern Europe. Product availability includes:
Coming soon:
Introducing console.vultr.com: The new Vultr customer portal is now available at console.vultr.com, providing a modern, enterprise-grade platform for managing infrastructure. With a new design and built-in enterprise IAM capabilities (policies, roles, groups, service accounts, and organization management), Vultr's console delivers improved usability, scalability, and control for teams of all sizes.
Archival Object Storage Tier Released: Vultr now offers a new, low-cost storage tier ($0.006/GB-month) designed for infrequently accessed data. This includes automatic lifecycle management and on-demand retrieval, helping optimize storage costs without sacrificing accessibility. See the documentation.
Vultr Serverless Inference Improvements: Vultr's Serverless Inference has been upgraded to support the latest generation of models with enhanced performance. Improvements include KV caching, expanded tool-calling support, and overall throughput gains. In addition, billing has transitioned to a usage-based model, with per-model input and output token pricing for greater flexibility and cost efficiency.
NVIDIA Build Partnership: Vultr has partnered with NVIDIA through the NVIDIA Build program. Vultr's serverless inference models are now accessible via the NVIDIA portal (build.nvidia.com), allowing customers to use a centralized NVIDIA Developer Program account linked to their Vultr account with a Serverless Inference subscription.
Load Balancer Scaling Fix: We have resolved an issue where load balancers could scale down to zero nodes, improving reliability and ensuring consistent availability.
UDP Support for Load Balancers: Load Balancers now support UDP traffic, enabling use cases such as real-time applications, gaming, streaming, and DNS services alongside existing TCP support.
OpenClaw One-Click App: Introducing the OpenClaw One-Click App, providing an easy way to consume our Serverless Inference offering through a popular agentic workflow. This enables faster setup and streamlined deployment for AI-driven applications. How to guide
Public API for Support Tickets: Introduced a new public API for support tickets, enabling customers to create, view, and manage tickets programmatically. This update removes the dependency on the UI for ticket workflows and allows support operations to be fully integrated into automated systems and tooling.
NAT Gateways Open-Source Tooling Support: NAT Gateways now include expanded open-source tooling support, adding key libraries such as Terraform. This enables customers to provision, manage, and automate NAT Gateway resources using infrastructure-as-code workflows and existing open-source ecosystems. Documentation
Bootable Block Storage Updates: Expanded support for bootable block storage, enabling greater flexibility and automation across below workflows. Documentation
These updates improve portability, disaster recovery options, and infrastructure-as-code support for block storage–backed deployments.
Location-Specific Pricing in List Plans API: The List Plans API response now includes location-specific pricing, allowing customers to query and compare plan costs on a per-location basis. This enhancement provides greater pricing transparency and enables more precise deployment and cost-optimization decisions.
VPS Plan Deprecation ("-sc1"): VPS plans ending in "-sc1" have been deprecated. No action is required for existing customers currently using these plans. New orders placed with "-sc1" plans will automatically be converted to the corresponding standard plan with location-specific pricing.
cPanel on AlmaLinux Added to Marketplace: A new cPanel Marketplace application based on AlmaLinux is now available. This new offering provides a stable, RHEL-compatible OS option for cPanel deployments, giving customers more flexibility and long-term support for hosting and management workloads.
Ubuntu 25.04 End of Life: Ubuntu 25.04 (Plucky Puffin) will reach End of Life on January 15, 2026. After this date, it will no longer receive security updates or maintenance. Customers are encouraged to plan upgrades to a supported Ubuntu LTS or interim release before the EOL date to maintain security and stability.
Fedora 44 Beta Available: Fedora 44 Beta is now available, giving customers early access to the latest features and updates.
FreeBSD 14.x Updated to 14.4 and Debian 13.x Updated to 13.4: These images have been updated to the latest point release with improvements and fixes.
Ubuntu NVIDIA Driver Compatibility Update: Ubuntu GPU offerings have been updated to use NVIDIA driver pinning packages, ensuring compatibility with NVIDIA's reworked driver packaging. This change improves driver stability and prevents unintended upgrades or mismatches across CUDA and GPU-enabled workloads.
NVIDIA GPU ECC Memory Enabled by Default: NVIDIA GPU ECC memory is now enabled by default across supported GPU offerings, following NVIDIA's recommended best practices. This change improves reliability and data integrity for compute-intensive and long-running GPU workloads without requiring manual configuration.
AppRepo GPG Key Reissued (SHA-256): The AppRepo GPG key expired on February 1 and has been reissued using SHA-256. Customers running images that rely on the previous key will need to update to the new GPG key to continue receiving AppRepo updates without interruption.
Maintenance Notifications Now Include Calendar Invites: Maintenance notification emails now include an .ics calendar attachment, allowing customers to easily add maintenance windows to their calendars. If the schedule changes, the event updates automatically without creating duplicates.
growpart and/or extend2fs.kimi-k2-instruct.PostgreSQL 13 EOL: PostgreSQL 13 has reached End of Life (EOL) and is no longer receiving security patches or community support. Customers should begin planning to upgrade their PostgreSQL 13 databases to a supported release. Use the Vultr control panel or API to deploy a new managed database on PostgreSQL 14+ and migrate your data. For upgrade guidance, click here.
Support for Additional API Keys: You can now generate additional API keys for your Vultr user. All API keys also now support an optional expiration date and name for convenient identification after they are generated. As always, API keys are displayed only once immediately after generation. Learn More.
DBaaS CA Certificate API Customers can now retrieve their account’s CA certificate for Managed Databases directly through the Vultr API. This CA certificate enables applications and services to establish secure, trusted TLS connections to Vultr’s Managed Databases.
Snapshot ID Display for VPS & Bare Metal Deploy: The Vultr API now retains and displays the snapshot ID used when deploying Virtual Private Servers (VPS) and Bare Metal (BM) instances. This allows customers to confirm which snapshot was used to create an instance. In addition, snapshot IDs can be retrieved programmatically for auditing, monitoring, or orchestration workflows.
Debian 13 ‘Trixie’ Now Available: The latest version of Debian has been added to the Vultr OS offering. Highlights include Linux 6.12 LTS with expanded hardware support,updated desktops (GNOME 48, KDE Plasma 6, Xfce 4.20), new “Ceratopsian” theme, improved installer with HTTP Boot, tmpfs default for /tmp, and APT 3.0 with modernized dependency resolution. Wayland is now default for GNOME, offering better multimedia and NVIDIA support.
We’re excited to announce that Identity and Access Management (IAM) and Organizations support is on the way! What to Expect:
These updates bring enterprise-grade controls to Vultr, making it easier for both individuals and businesses to securely scale their operations. Stay tuned — more details coming soon!
VKE Node Pools Now Support User Data: Vultr's Terraform provider now supports user_data on the vultr_kubernetes_node_pools resource. You can now specify user data per node pool, giving users more control and customization at launch. Learn More.
Account Logs are Now Available: Logs provide detailed records of user actions across your account, capturing events such as logins, Vultr Console interactions, and API requests. Learn more.
Ubuntu 25.04 BM Bug Fixes: Improved support for Ubuntu 25.04 across Bare Metal hardware, ensuring smoother deployment and stability.
Global Load Balancer Experience Improvements: A redesigned, modernized interface for automatically routing traffic to the nearest server worldwide, making global traffic management easier and more intuitive. Check it Out.
High Performance Block, Object, and Filesystem Storage in Seattle and Atlanta: Seattle and Atlanta are key hubs for current and future GPU deployments. High-performance storage located near customer GPU workloads ensures optimal performance for compute-intensive applications.
NVIDIA HGX B200 or H100 GPU Plans Available: Deploy Cloud GPU plans featuring 8x NVIDIA HGX B200 or H100 GPUs. These full-system virtual machines combine bare-metal performance with cloud agility, featuring: 8x GPUs in passthrough mode for maximum performance and full GPU fabric networking for inference and training workloads. Deploy Now.
Rapid AI deployment with Vultr Open Cluster Manager: Vultr Open Cluster Manager is now available as a one-click-deploy app on the Vultr Marketplace. Powered by open-source technologies, Vultr Open Cluster Manager simplifies the process for getting AI workloads running on Vultr Cloud GPU clusters. Learn more.