> For the complete documentation index, see [llms.txt](https://docs.stg.ai.neevcloud.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.stg.ai.neevcloud.com/release-notes/apr-2026-release.md). # April 2026 Release **Release Type:** Patch Release\ **Release Version:** v0.3.2 *** ## Overview The April 2026 release brings a significant pricing reduction on Network Storage, real-time streaming in the Inferencing Playground, and an important security fix. This release reflects our ongoing commitment to making NeevCloud more transparent, cost-efficient, and secure for every team building on the platform. *** ## What's New ### Storage — Network Storage Pricing Update We have significantly reduced the Network Storage billing rate from **Rs. 9/GB/month to Rs. 2/GB/month** — a reduction of over 77%. This change is reflected consistently across billing, the platform UI, and all invoices generated from this release date onwards. If you are running storage-heavy workloads — such as model checkpoints, training datasets, or long-running deployment volumes — you will see a direct and immediate reduction in your monthly bill with no action required on your end. This pricing update applies to all active Network Storage volumes across your account. ### AI Inference — Streaming Support Streaming is now supported across both the Inferencing Playground and the Model API endpoint, giving you real-time token-by-token output as the model generates its reply. Previously, responses were returned only after generation was fully complete. This made it difficult to evaluate model behaviour on longer outputs, debug prompts in real time, or build responsive end-user experiences. With streaming enabled, output is delivered incrementally as it is generated — whether you are testing in the Playground or integrating directly via the Model API. To use streaming in the Playground, toggle the Streaming option in the interface before submitting your prompt. For the Model API, refer to the streaming section in the API reference. *** ## Bug Fixes and Improvements ### GPU — Resource Scaling Fix **Fixed:** CPU and Memory resources now scale proportionally when a workload is allocated multiple GPUs. If you were running multi-GPU workloads, you may have noticed degraded performance or resource shortfalls. This was caused by compute resources not scaling alongside GPU count. This has now been corrected — when you allocate more GPUs to a workload, the associated CPU and Memory scale accordingly, ensuring your workloads always have the full resources they need. ### GPU Availability — "See All" Removed The "See All" option in the GPU availability view has been removed. GPU availability is now shown directly in the GPU list, giving you a clear and consolidated view without needing an additional step to expand the full list. ### Billing — Reserved Plan Savings Accuracy Fix **Fixed:** The savings calculation displayed on Reserved Pricing Plans now reflects the correct amount. The previous display was showing incorrect savings figures, which made it difficult to accurately assess the value of committing to a reserved plan versus on-demand pricing. The corrected calculation now accurately shows the savings you unlock by choosing a reserved plan, helping you make better-informed decisions about your compute commitment. *** ## Security Fixes ### gRPC-Go Authorization Bypass Mitigation **Fixed:** A potential authorization bypass vulnerability related to a missing leading slash in the `:path` header in gRPC-Go has been mitigated. This issue could, under specific conditions, allow requests to bypass authorization checks if the `:path` value did not include a leading slash as expected by the routing and authorization layer. The fix ensures that path validation is enforced correctly, closing the potential bypass vector. No customer action is required — the fix is applied at the infrastructure level. If you have any questions about this fix or its impact on your workloads, please reach out to the NeevCloud support team. *** ## Known Issues * Storage volume size cannot be changed after creation. If you need additional capacity, create a new storage volume. * GPU startup time may vary based on region and current availability. * GPU resizing and live migration are not currently supported. * Payments are processed in USD. Approximate local currency displays including INR are shown in the UI — full local currency billing support is planned for upcoming releases. *** ## Notes The Network Storage pricing reduction in this release is a direct result of infrastructure cost optimisations we have been working on over the past quarter. We are committed to passing these savings on to you as we scale. We continue to prioritise stability, security, and pricing transparency in every release. Keep sharing your feedback — it shapes what we build next.