> 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/july-2026-release.md). # July 2026 Release **Release Type:** Major Feature Release\ **Release Version:** v0.7.0 *** ## Overview The July 2026 release brings **GPU Virtual Machines** to NeevCloud, along with a few improvements to **Agentic Studio** and more flexibility for deploying your own models. This release includes: * GPU Virtual Machines, with usage metrics and a ready-to-go GPU image * A **10 TB storage increase** across GPU AI Service and VM Service, with multi-disk support * Bring-your-own-container-image support for Sandboxes, plus public preview links and simpler SSH access * New egress controls for Sandboxes and Agents * Broader support for multimodal and generative models on dedicated deployments * New promotional credits and clearer wallet balances * New agent templates in the Agentic Studio catalogue Together, these updates give you more compute options, more model flexibility, and clearer visibility into your usage and spend. *** ## What's New ### GPU Virtual Machines *(New)* You can now launch GPU-powered virtual machines on NeevCloud, alongside our existing CPU VMs. * A pre-built Ubuntu OS image, pre-loaded with CUDA and PyTorch, so you can get started immediately * **Usage metrics** for CPU, RAM, and disk, right in the platform **Coming soon:** Pre-built AI templates for GPU VMs (for now, you'll need to install your own dependencies). *** ### Storage *(New, Common)* Network storage volumes have been expanded across the platform. * **Up to 10 TB of storage**, with support for multiple disks, now available for both **GPU AI Service** and **VM Service** ### SSH Key Management *(New, Unified)* SSH keys can now be managed from a single place, and used across different products and services. * Create, view, and delete SSH keys from one consistent place * No more managing multiple keys across different products *** ### Agentic Studio — Sandboxes Sandboxes get several upgrades this release, giving you more control over your environment and easier ways to work with files and preview your work. * **Bring your own container image** to run your sandboxes exactly the way you want, with built-in safeguards to keep image sizes reasonable * **Public preview links** for anything your sandbox exposes, so you can share a running preview instantly * **Simpler, keyless SSH access** — connect to your sandbox without managing keys yourself * **Reconnect without losing your place** — terminal sessions now survive a disconnect * More ways to manage files inside your sandbox, plus **real-time notifications** when files change * **Full memory and workspace snapshots** are now supported. When you restore, fork, or resume from a snapshot, your process continues exactly from the point where the snapshot was taken. Paused sandboxes are kept for **14 days** before being deleted. **Coming soon:** Sandbox startup times moving from seconds to milliseconds, and more control over idle sandboxes. ### Agentic Studio — Agents *(Beta, updated)* * **Hermes** joins the agent template catalogue * **New metrics dashboard** to track how your agents are performing **Coming soon:** Full pause-and-resume support for agents (currently available for Sandboxes only). **Note:** Agents remain in Beta. *** ### Network Egress Controls *(New, Common)* We've added egress policies that let you control internet access for your **Sandboxes** and **Agents**, including the option to enable unrestricted outbound internet access. *** ### Developer Tools * Manage sandboxes right from the command line — checkout, fork, and restore in one command * Cleaner, easier-to-read sandbox listings in the CLI * Updated JavaScript and Python SDKs * New documentation makes it easier to build sandbox functionality into your own tools * Command output now streams live by default * Stronger security with separated access tokens for platform management vs. running your sandboxes *** ### AI Inference — Dedicated Model Deployment * Deploy models that understand both **images and text** (including gemma-4 and Qwen models) as dedicated endpoints * New support for models that **generate** images, video, and audio — not just text * **Run your own benchmarks on demand**, right from a new benchmarking view in the platform * Better visibility into how your dedicated models are performing *** ### Billing & Credits * **New promotional credits**, with clearer visibility into your available balance * Improved accuracy and consistency across your billing history * Wallet balances now shown more clearly, including your total available credit per product *** ## Known Issues ### GPU Virtual Machines * Pre-built AI templates aren't available yet for GPU VMs — you'll need to install dependencies manually for now. ### AI Inference — Dedicated Model Deployment * When manually choosing the number of GPUs for a deployment, only certain counts (1, 2, 4, 8, and so on) are supported today. Choosing an unsupported number can cause the deployment to fail. We recommend using the automatic configuration option, which picks a supported count for you. * The benchmarking view can currently overlap with the model details view. A cleaner layout is coming soon. ### Agentic Studio — Sandboxes * Forking a sandbox through our SDK isn't working correctly right now (forking through the CLI is unaffected). A fix is planned for the next patch release. * Writing a file larger than your sandbox's allocated storage doesn't fail right away — you'll only see an error after the sandbox stops due to exceeding its storage limit. We're working on catching this earlier. ### Agentic Studio — Agents * Full pause-and-resume (memory snapshot) isn't available for Agents yet. * You may occasionally notice a short delay before the Agents view refreshes automatically. * The OpenCode template icon will get a visual refresh in an upcoming release. *** ## Notes This release brings GPU compute and expanded storage to the platform, and rounds out Agentic Studio with bring-your-own images, richer file handling, and broader SDK support — alongside expanded multimodal and generative model options for dedicated deployments. Sandboxes and Agents remain in Beta as we keep shipping improvements every release. As always, your feedback shapes what we build next. If you run into issues or have suggestions, reach out via a support ticket or contact your NeevCloud account team. *** ## Understanding Beta Features marked as *(Beta)* are fully available to all users and ready to use, but are still being actively improved. Here is what to expect: * **Core functionality works.** Beta features are not experimental — they are production-capable, but have not yet met all the stability and completeness thresholds required for General Availability (GA). * **Frequent improvements.** You can expect updates to performance, reliability, and functionality in every release cycle. * **Some gaps remain.** Certain views or flows may not be fully complete yet. Known gaps are listed in the Known Issues section above. * **APIs may evolve.** Core endpoints are stable, but some parameters or response formats may be refined before GA. * **No full SLA coverage.** Beta features are not covered by the same uptime commitments as GA features. * **Your feedback matters.** Beta is a collaborative phase — the issues you raise and the workflows you share directly shape what we prioritise next. Please reach out. Beta features graduate to GA when they meet our internal quality, performance, and completeness thresholds. GA graduation is always announced in the release notes.