Releasesby GlyphDeepSeek V4 Flash now runs updated weights on AI Gateway
DeepSeek V4 Flash now runs on updated weights by default on AI Gateway, with notably stronger agentic capabilities. On Terminal-Bench, it scores 82.7, up 25.8 points from 56.9 in the April preview. Requests to deepseek/deepseek-v4-flash pick up the new weights automatically, with no change to the model ID or your code.
Also shipped Jul 31, 2026Vercel in July 2026 →
Web terminal on serverless GPU compute (AI Runtime) is in Public Preview
The web terminal in notebooks connected to serverless GPU compute (AI Runtime) is now in Public Preview . Use it to run shell commands, monitor GPU usage with nvidia-smi , install libraries, and manage files directly on the compute. This requires serverless GPU environment version 5 or above. See Run shell commands in Databricks web terminal .
Packaged clean rooms are now generally available
Packaged clean rooms are now generally available. In a packaged clean room, a provider supplies notebooks, JARs, and data that a consumer runs against their own data without seeing the provider's code or data assets. See Packaged clean rooms .
OpenAI connector (Beta)
The OpenAI connector is now available in Lakeflow Connect in Beta . Workspace admins can control access to this feature from the Previews page. The connector uses the OpenAI Organization Admin API to ingest organization administration data, such as users, projects, API keys, usage, costs, and audit logs, from OpenAI into Databricks.
[In preview] Public Preview: Azure SQL updates for late-July
In late-July 2026, the following update and enhancement was made to Azure SQL: Regex-based dynamic data masking (DDM) extends the built-in DDM feature with pattern‑based masking to hide sensitive data like emails, phone numbers, and identifiers, centrally
Upcoming August 2026 model deprecations in GitHub Copilot
We will deprecate the following models across all GitHub Copilot experiences (including Copilot Chat, inline edits, ask and agent modes, and code completions), on September 1st, 2026: Model Deprecation date Suggested alternative Gemini 3.1 Pro 9-1-2026 Gemini 3.6 Flash Claude Opus 4.5 9-1-2026 Claude Opus 4.7, Claude Opus 4.8, or Claude Opus 5 Claude Opus...
Gemini 2.5 Pro and Gemini 3 Flash deprecated
As of today, July 31, 2026, we have deprecated the following models across all GitHub Copilot experiences (including Copilot Chat, inline edits, ask and agent modes, and code completions).
More from AI SDK and AI GatewayEvery AI SDK and AI Gateway release →
AI Gateway now supports team and project spend budgets
AI Gateway spend budgets can now be scoped to a team or a project, in addition to individual API keys. Set a dollar limit on a scope, and the gateway meters spend against it and stops further requests once the limit is reached, until the budget resets or you raise it. Scopes A budget attaches to one of three scopes.
AI Gateway logs now have a dedicated page
AI Gateway now has a dedicated Logs page listing every request your team sends through the gateway, newest first, with cost, token counts, duration, and the model, provider, and region that served it. Open any request to see how it was routed. It's available at the team scope and per project.
Expanded search for workflow runs in Vercel Observability
The runs list for Vercel Workflows in Observability now has structured search, giving you more ways to find and investigate runs. Structured filters: Filter runs by workflow, environment, deployment ID, region , or any custom run attribute. Paste a run ID to pull up that exact run.
10x more capacity for Laguna S 2.1 on AI Gateway
Laguna S 2.1 from Poolside now has 10x more capacity on AI Gateway. The increase applies to the paid version, poolside/laguna-s-2.1 , and the free version, poolside/laguna-s-2.1-free , so you can send far more requests, good for high-volume agentic coding and long-running tasks.
AI Gateway adds unified fast mode support
AI Gateway has a unified fast mode abstraction, now in beta. You can now request fast mode the same way for every model on AI Gateway. Set speed to fast , and the gateway serves the fast tier when it's available and falls back to standard speed when it isn't. Fast mode trades a higher per-token cost for lower latency or higher throughput.
Run multiple isolated agents in a single Sandbox
The @vercel/sandbox SDK now supports multiple Linux users and groups, so you can run agents side by side in a single Sandbox. Each agent runs as its own user with a private home directory. A group opens a shared workspace when they need to collaborate. This makes multi-agent systems easier to build.
MiniMax H3 now available on AI Gateway
MiniMax H3 is now available on AI Gateway. H3 generates 2K video from a text prompt, a starting image, a pair of first and last frames, or reference material. Alongside text-to-video and first-frame image-to-video, the model supports first-to-last keyframe transitions and multimodal reference-to-video, conditioning a generation on reference images, video...
Inkling Small from Thinking Machines is now available on AI Gateway
Inkling Small from Thinking Machines is now available on AI Gateway. Inkling Small reaches performance comparable to the larger Inkling model at about a quarter of the size, using much less compute per task. It is a broad generalist with native reasoning over audio and images, and it holds up well on reasoning, agentic coding, and tool use.