AI Pulse

<h2>What's changed</h2> <ul> <li>Added Claude Mods: plugins may now modify deeper behavior</li> <li>Added You should know, an opt-in plugin where a side agent watches your back and flags things you or Claude might miss. Turn it on with <code>/plugin enable cc-plugin-you-should-know@builtin</code> (for first-party sessions with telemetry on)</li> <li>Added an <code>n:<text></code> filter to the agents view that matches session names and tasks; a filter now shows matches in collapsed sections and

<h3>Misc Changes</h3> <ul> <li>Always use generateBuildId if set explicitly: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5565264262" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99147" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99147/hovercard" href="https://github.com/vercel/next.js/pull/99147">#99147</a></li> <li>Set tracer library version: <a class="issue-link js-issue-link"

Vercel and Microsoft AI (MAI) have partnered to make MAI models available on AI Gateway. AI Gateway is one of a limited number of platforms offering access to MAI models. MAI's focus on security and zero data retention (ZDR) align with Vercel's goal of giving AI Gateway users control over their data and transparency in how it is used for training. MAI models on AI Gateway MAI's latest audio models are now supported on AI Gateway. MAI-Voice-2.1 and MAI-Voice-2.1-Flash generate speech, while MAI-T

<h3>Misc Changes</h3> <ul> <li>Scope response cache keys to their source route: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5649279227" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99482" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99482/hovercard" href="https://github.com/vercel/next.js/pull/99482">#99482</a></li> <li>Match Next data paths case-sensitively: <a class="issue-link

Vercel Agent can install private dependencies from npm and custom registries using credentials stored as shared environment variables on Vercel. npm , pnpm , and classic Yarn running in Agent sessions authenticate as they do in Vercel builds. To get started, add a shared environment variable for Development or Preview: Use NPM_TOKEN for private packages hosted on registry.npmjs.org . Use NPM_RC to configure custom or multiple registries. Vercel Agent reads only team-shared variables, not project

<h2>0.0.17 (2026-09-30)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/google-cloud-sdk-v0.0.16...google-cloud-sdk-v0.0.17">google-cloud-sdk-v0.0.16...google-cloud-sdk-v0.0.17</a></p>

<h2>0.5.2 (2026-09-30)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/foundry-sdk-v0.5.1...foundry-sdk-v0.5.2">foundry-sdk-v0.5.1...foundry-sdk-v0.5.2</a></p>

<h2>0.34.2 (2026-09-30)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/bedrock-sdk-v0.34.1...bedrock-sdk-v0.34.2">bedrock-sdk-v0.34.1...bedrock-sdk-v0.34.2</a></p> <h3>Chores</h3> <ul> <li><strong>client:</strong> deprecate sonnet 4.5 (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/d72cfaf6a551c75d54fad0538fb5fd65fb2346d1">d72cfaf</a>)</li> </ul>

<h2>0.20.2 (2026-09-30)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/vertex-sdk-v0.20.1...vertex-sdk-v0.20.2">vertex-sdk-v0.20.1...vertex-sdk-v0.20.2</a></p> <h3>Chores</h3> <ul> <li><strong>client:</strong> deprecate sonnet 4.5 (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/d72cfaf6a551c75d54fad0538fb5fd65fb2346d1">d72cfaf</a>)</li> </ul>

<h2>0.131.0 (2026-09-30)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/sdk-v0.130.0...sdk-v0.131.0">sdk-v0.130.0...sdk-v0.131.0</a></p> <h3>Features</h3> <ul> <li><strong>api:</strong> add list spend limits endpoint (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/cfb78c8dde70b5f9ea07bf2aebd3651282f2c9a8">cfb78c8</a>)</li> </ul> <h3>Chores</h3> <ul> <li><strong>client:</strong> deprecate sonnet 4.5 (<a href="https://github.

<h2><a href="https://github.com/googleapis/python-genai/compare/v2.25.0...v2.26.0">2.26.0</a> (2026-09-30)</h2> <h3>Features</h3> <ul> <li>Add gcs_metrics_uri for exporting tuning job metrics. (<a href="https://github.com/googleapis/python-genai/commit/b6535e822a6a984bec43056551b2788cd61509f0">b6535e8</a>)</li> <li>Add Gemini 3.8 Flash TTS and Gemini 3.8 Flash Lite TTS models to SDKs (<a href="https://github.com/googleapis/python-genai/commit/0cad6ca7cfc4dce6f58784f546603d8cc8a5f921">0cad6ca</a>

Browserbase Search and Fetch are now available through AI Gateway . Give your models access to current information, find relevant pages, and retrieve their contents, alongside Browserbase's existing AI SDK and Vercel Marketplace integrations. AI Gateway lets you use Browserbase's tools across model providers with one API key. Add web search and page retrieval to any model that supports tool calling, and keep the same tools when you switch models. These helpers are available in AI SDK 7.0.116 and

CDN Requests is now the name for the usage metric previously called Edge Requests. This is a naming change only: pricing, limits, and how usage is measured are all unchanged. The new name appears on the usage dashboard, the Observability Events chart, usage notifications, and invoice line items. It reflects that these are requests served by Vercel's CDN. Metric IDs and product aliases are unchanged, so saved queries keep working. In the billing charges API, ServiceName now returns the new name.

<h2>What's changed</h2> <ul> <li>Added a count such as "2 of 5" to the permission prompt when several permission requests stack up</li> <li>Added mouse support for the "N more" rows of lists in fullscreen mode: click one to jump to that end of the list, with hover and pressed states</li> <li>Fixed several Claude Code processes and IDE extensions each opening a login browser when gcpAuthRefresh or awsAuthRefresh credentials expire</li> <li>Fixed <code>claude --resume</code> and <code>--continue</

<h2>0.0.16 (2026-09-30)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/google-cloud-sdk-v0.0.15...google-cloud-sdk-v0.0.16">google-cloud-sdk-v0.0.15...google-cloud-sdk-v0.0.16</a></p>

<h2>0.5.1 (2026-09-30)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/foundry-sdk-v0.5.0...foundry-sdk-v0.5.1">foundry-sdk-v0.5.0...foundry-sdk-v0.5.1</a></p>

<h2>0.34.1 (2026-09-30)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/bedrock-sdk-v0.34.0...bedrock-sdk-v0.34.1">bedrock-sdk-v0.34.0...bedrock-sdk-v0.34.1</a></p>

<h2>0.20.1 (2026-09-30)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/vertex-sdk-v0.20.0...vertex-sdk-v0.20.1">vertex-sdk-v0.20.0...vertex-sdk-v0.20.1</a></p>

<h2>0.130.0 (2026-09-30)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/sdk-v0.129.0...sdk-v0.130.0">sdk-v0.129.0...sdk-v0.130.0</a></p> <h3>Features</h3> <ul> <li><strong>api:</strong> add a refusal stop reason and stop_details to Managed Agents session idle events (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/5cd6910a0e74923f1bef35f5368f9c2cee410058">5cd6910</a>)</li> <li><strong>api:</strong> add Claude Enterprise anal

<p>This release contains security fixes for the following advisories:</p> <p>High:</p> <ul> <li><a href="https://github.com/vercel/next.js/security/advisories/GHSA-cjq9-62q9-8jv4">Server-Side Request Forgery in Image Optimization</a></li> </ul> <p>Medium:</p> <ul> <li><a href="https://github.com/vercel/next.js/security/advisories/GHSA-f87g-xv8r-7p7x">Information disclosure in Next.js App Router metadata image routes via dynamicParams bypass</a></li> <li><a href="https://github.com/vercel/next.js

<p>This release contains security fixes for the following advisories:</p> <p>Medium:</p> <ul> <li><a href="https://github.com/vercel/next.js/security/advisories/GHSA-f87g-xv8r-7p7x">Information disclosure in Next.js App Router metadata image routes via dynamicParams bypass</a></li> <li><a href="https://github.com/vercel/next.js/security/advisories/GHSA-4jqv-mc3x-m676">Cache poisoning of SSG and ISR pages in self-hosted Next.js applications</a></li> <li><a href="https://github.com/vercel/next.js/

Vercel CDN no longer caches origin responses when Vary includes Cookie . Vary tells a cache which request headers may change the response. Because cookies often contain visitor-specific values, varying by Cookie can create many cache entries that are rarely reused. The response is still served normally but isn't stored for future requests. You can identify these responses by checking the x-vercel-cache header and Runtime Logs . The header is set to MISS , and Logs show vary_key_denied:cookie as

Ling 3.1 Flash from InclusionAI is now available on AI Gateway. The model is free to use through October 13, 2026. Ling 3.1 Flash is a hybrid reasoning language model with 560B total parameters and 25B active per token. It has a 262K-token context window on AI Gateway. The model is designed for coding, multi-step analysis, and agents that use tools, including workflows involving long documents, code, and extended task histories. To try Ling 3.1 Flash, use inclusionai/ling-3.1-flash or inclusiona

Vercel Sandbox now supports Secure Compute , connecting sandboxes to a team's dedicated network. Public-internet traffic exits through the network's static IPs, and sandboxes can reach private resources in an AWS VPC through VPC peering. You can attach a sandbox to your network in your code or via the CLI. In your code, pass an existing Secure Compute network ID when creating a sandbox: In the CLI, append --network-id and your network id value when you run the sandbox create command: Existing sa

<h3>Misc Changes</h3> <ul> <li>[create-next-app] Enable Cache Components by default in create-next-app: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5635062996" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99436" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99436/hovercard" href="https://github.com/vercel/next.js/pull/99436">#99436</a></li> <li>Include async module evaluation in cl

Vercel Connect now lets you submit your own service to the connectors directory , so you no longer need to wait for Vercel to build a connector for it. Submitting a service takes four steps: Describe your service. Add one or more connection methods, using OAuth or an API key. Verify each OAuth method by requesting a real token through a test connector. Send the submission for review. Vercel reviews each submission before publishing it to the directory. Get started from Submit a Service in the da

<h2>What's changed</h2> <ul> <li>Added <code>CLAUDE_CODE_DISABLE_WEB_FETCH</code> environment variable to turn off the WebFetch tool</li> <li>Added <code>claude --desktop</code> to open the Claude desktop app on the current directory, or on a session with <code>--continue</code> / <code>--resume <id></code></li> <li>Added <code>claude plugin configure <plugin></code> to show a plugin's options and which are unset, or save new values read from stdin with <code>--values-stdin</code></li> <li>Added

You and your coding agents can now search a project’s trace spans from the terminal with vercel traces search . Each span represents one step in a request, helping you investigate errors, latency, and security-related behavior without opening the dashboard. By default, vercel traces search returns up to 100 spans from the last hour, newest first, so you can search without first finding a request or trace ID. Filter results by environment, service, span name, status, deployment ID, request ID, tr

<div class="markdown-alert markdown-alert-note"><p class="markdown-alert-title"><svg data-component="Octicon" class="octicon octicon-info mr-2" viewBox="0 0 16 16" version="1.1" width="16" height="16" aria-hidden="true"><path d="M0 8a8 8 0 1 1 16 0A8 8 0 0 1 0 8Zm8-6.5a6.5 6.5 0 1 0 0 13 6.5 6.5 0 0 0 0-13ZM6.5 7.75A.75.75 0 0 1 7.25 7h1a.75.75 0 0 1 .75.75v2.75h.25a.75.75 0 0 1 0 1.5h-2a.75.75 0 0 1 0-1.5h.25v-2h-.25a.75.75 0 0 1-.75-.75ZM8 6a1 1 0 1 1 0-2 1 1 0 0 1 0 2Z"></path></svg>Note</p><

GPT-6.1 Sol from OpenAI is now available on AI Gateway. It improves on GPT-6 Sol for coding, computer use, and professional work, including reading complex documents and carrying out multi-step workflows. The model is suited to agents that need to debug code, work through business tasks, or extract answers from PDFs with tables and charts. It also improves factual accuracy on difficult questions. Its standard input and output pricing is lower than GPT-6 Astra's, while cheaper cached input makes

<h3>Misc Changes</h3> <ul> <li>[test] Stabilize middleware revalidation test: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5620494432" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99378" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99378/hovercard" href="https://github.com/vercel/next.js/pull/99378">#99378</a></li> <li>test: scope loader build dependency update to Turbopack: <a cla

<img src="https://storage.googleapis.com/gweb-uniblog-publish-prod/images/futurevisionxprize_social.max-600x600.format-webp.webp">Watch the winning trailer from the Future Vision XPRIZE, The Gifted.

<h2>0.0.15 (2026-09-28)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/google-cloud-sdk-v0.0.14...google-cloud-sdk-v0.0.15">google-cloud-sdk-v0.0.14...google-cloud-sdk-v0.0.15</a></p>

<h2>0.7.5 (2026-09-28)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/aws-sdk-v0.7.4...aws-sdk-v0.7.5">aws-sdk-v0.7.4...aws-sdk-v0.7.5</a></p>

<h2>0.5.0 (2026-09-28)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/foundry-sdk-v0.4.9...foundry-sdk-v0.5.0">foundry-sdk-v0.4.9...foundry-sdk-v0.5.0</a></p> <h3>Features</h3> <ul> <li><strong>api:</strong> add claude-sonnet-5-5 (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/4e7736625fd23a5607800e8de4aa3c37daa74a07">4e77366</a>)</li> </ul>

<h2>0.34.0 (2026-09-28)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/bedrock-sdk-v0.33.8...bedrock-sdk-v0.34.0">bedrock-sdk-v0.33.8...bedrock-sdk-v0.34.0</a></p> <h3>Features</h3> <ul> <li><strong>api:</strong> add claude-sonnet-5-5 (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/4e7736625fd23a5607800e8de4aa3c37daa74a07">4e77366</a>)</li> </ul>

<h2>0.20.0 (2026-09-28)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/vertex-sdk-v0.19.11...vertex-sdk-v0.20.0">vertex-sdk-v0.19.11...vertex-sdk-v0.20.0</a></p> <h3>Features</h3> <ul> <li><strong>api:</strong> add claude-sonnet-5-5 (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/4e7736625fd23a5607800e8de4aa3c37daa74a07">4e77366</a>)</li> </ul>

<h2>0.129.0 (2026-09-28)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/sdk-v0.128.0...sdk-v0.129.0">sdk-v0.128.0...sdk-v0.129.0</a></p> <h3>Features</h3> <ul> <li><strong>api:</strong> add between_tools thinking type (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/ab51075609a2d583dffdca24856f4214779d9e7f">ab51075</a>)</li> <li><strong>api:</strong> add claude-sonnet-5-5 (<a href="https://github.com/anthropics/anthropic-sdk

<h2>What's changed</h2> <ul> <li>Added Claude Sonnet 5.5 (<code>claude-sonnet-5-5</code>), now the default Sonnet model on the Anthropic API — 1M context, $2/$10 per Mtok with $0.20/Mtok cache reads</li> <li>Added a "Yes, but ask again next time" answer to auto mode's prompt before a read outside the working directories, so you can allow that one read and still be asked about later ones</li> <li>Added dollar amounts to the Claude apps gateway spend limit in <code>/usage</code> and the status lin

Claude Sonnet 5.5 from Anthropic is now available on AI Gateway as anthropic/claude-sonnet-5.5 . Sonnet 5.5 improves on Claude Sonnet 5 for well-scoped everyday work, from building features and fixing bugs to creating documents, slides, and spreadsheets. On coding tasks, Sonnet 5.5 carries changes through multiple steps and checks its work before reporting completion. It produces more polished documents, slides, and spreadsheets with fewer follow-up edits, and writes clearer explanations and rep

You can now discover domain names and check availability and pricing without signing in to Vercel or providing an access token. This works through the Vercel CLI and Domains Registrar API. With the Vercel CLI , search by keyword or domain fragment: With the Domains Registrar API , check up to 200 exact domain names in one request. The response shows whether each can be registered and, when available, its registration and renewal prices: Authentication is still required to buy or manage domains.

<h3>Misc Changes</h3> <ul> <li>Show security upgrade insights in DevTools: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5532013872" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99002" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99002/hovercard" href="https://github.com/vercel/next.js/pull/99002">#99002</a></li> <li>Use focused npm advisories and nudge only for ready upgrades: <a c

Ember-1 from Fireworks is now available on AI Gateway . Ember-1 is a research preview reasoning model built on Kimi K3 for coding and agentic workflows. Fireworks reports approximately 40% fewer generated tokens than Kimi K3 at comparable quality across its evaluations. For coding agents that make repeated model calls, shorter reasoning traces can reduce output costs and the amount of context carried into later steps. Ember-1 supports a 1M-token context window, text and image input, tool calling

<h3>Misc Changes</h3> <ul> <li>Explain prerelease security upgrade limits: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5581868278" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99230" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99230/hovercard" href="https://github.com/vercel/next.js/pull/99230">#99230</a></li> <li>Support latest agentic upgrades on prerelease channels: <a class="

<h3>Misc Changes</h3> <ul> <li>Avoid facade splits solely for export mangling: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5596176942" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99285" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99285/hovercard" href="https://github.com/vercel/next.js/pull/99285">#99285</a></li> <li>test: forward __NEXT_PARTIAL_PREFETCHING to test-deploy: <a cl

<h3>Misc Changes</h3> <ul> <li>Hoist resolved pathname into app page route match: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5588012546" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99259" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99259/hovercard" href="https://github.com/vercel/next.js/pull/99259">#99259</a></li> <li>docs: papercuts in image API reference: <a class="issue-lin

<h2>What's changed</h2> <ul> <li>Added <code>x-claude-code-prompt-id</code> to the gateway hint headers so LLM gateways can group the requests that serve one user prompt; opt in with <code>CLAUDE_CODE_GATEWAY_HINT_HEADERS=1</code></li> <li>Added <code>availableModelsMatch</code> managed setting: with <code>"exact"</code>, an <code>availableModels</code> entry allows only the model version it names, so new releases stay blocked until listed</li> <li>Added <code>deniedModels</code> managed setting

<h3>Misc Changes</h3> <ul> <li>Prepare <code>'use cache'</code> for static root-param tracking: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5583897776" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99239" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99239/hovercard" href="https://github.com/vercel/next.js/pull/99239">#99239</a></li> <li>[test] Adjust the prerender-encoding open fal

<h3>Misc Changes</h3> <ul> <li>Revert "Add experimental custom webpack support": <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5581647558" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99227" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99227/hovercard" href="https://github.com/vercel/next.js/pull/99227">#99227</a></li> <li>test: stabilize lazy dynamic import error overlay: <a class=

You can now push container images from GitHub Actions to Vercel Container Registry (VCR) without storing long-lived registry credentials. The new vercel/vcr-action/login action authenticates your workflow using GitHub OIDC. It exchanges the workflow’s OIDC token for a short-lived Vercel access token, then uses that token to log in to vcr.vercel.com . When the job ends, the action logs out and revokes the Vercel token. To start: Create an OIDC policy on your Vercel team that matches the GitHub re

In seven months, the skills.sh registry grew to one million agent skills and recorded nearly 280 million installs. A skill gives an AI agent reusable instructions for a particular job. Agents are capable but generic; they can do many jobs, but they don't know how a specific person, team, or company does them. A skill supplies that missing context and can be as simple as a file written in ordinary language. Anthropic introduced Agent Skills in October 2025, and Vercel launched the skills.sh regis

You can now run Redis commands, browse keys, and inspect values directly in the Vercel dashboard using the Marketplace database browser . Both the Redis and Upstash for Redis integrations in Vercel Marketplace are supported, so you no longer need redis-cli or a separate database UI. From your project, open Infrastructure > Storage and select a Redis database. The database page now includes two new tabs: CLI: Run Redis commands and inspect the results. Run multiple commands by separating them wit

Vercel Sandbox observability now includes memory usage data. You can access sandbox memory usage data in the dashboard and through the CLI via the vercel metrics command. Sandbox observability memory in the dashboard The Memory Usage card reports average, P75, and P95 memory across your sandboxes, alongside the existing CPU usage and data transfer metrics on the sandbox overview page and the project and team-level pages. On the sandbox detail page, charts are designed to show how close a sandbox

Pixel Canary is now available on AI Gateway as stealth/pixel-canary . It's free for a limited time while in stealth. Pixel Canary is strong at coding, including building applications and refactoring existing code. It is well suited to frontend development and mobile app design, from responsive layouts and app screens to navigation and interactive components. On Next.js evals , Pixel Canary ties GPT 6 Astra (high) at a 90.3% baseline success rate, passing 28 of 31 tasks. With Next.js documentatio

<h3>Misc Changes</h3> <ul> <li>[test] Fix new deploy suites: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5569659496" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99162" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99162/hovercard" href="https://github.com/vercel/next.js/pull/99162">#99162</a></li> <li>Provide only the relevant context for agent upgrade: <a class="issue-link js-iss

<h2>What's changed</h2> <ul> <li>Added a <code>maxProseWidth</code> setting that caps the width of Claude's prose in wide terminals while tables and code blocks keep the full width</li> <li>Added a startup notice, and <code>/status</code> and <code>claude doctor</code> entries, listing telemetry variables in a project's settings files that were ignored or that turned telemetry off</li> <li>Added the <code>allowClaudeInChromeWithManagedMcp</code> managed setting to let <code>claude --chrome</code

<h3>Misc Changes</h3> <ul> <li>Turbopack: chunk each layout segment once in production: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5553878502" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99100" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99100/hovercard" href="https://github.com/vercel/next.js/pull/99100">#99100</a></li> <li>bundle-analyzer: exclude traces from output types: <a

<h3>Misc Changes</h3> <ul> <li>Tune Node server chunking for require overhead: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5559390190" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/99120" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/99120/hovercard" href="https://github.com/vercel/next.js/pull/99120">#99120</a></li> <li>Turbopack: fix another "ModuleId not found for ident" regressi

In 2022, we launched a private bug bounty program through HackerOne. Over the last several years, we've worked with HackerOne's VIP program to refine our process, targets, and scope, onboarding thousands of their best researchers and hardening security across our platform. We've translated those learnings into several public bounty programs: The $1 million React2Shell hacker challenge The $1 million Vercel Sandbox hacker challenge Our OSS bug bounty program Today, we are combining our private an

Klaviyo on Vercel Built a platform for shipping internal apps 512 builders and 356 live apps in two weeks Every app ships SSO-gated and private by default Builders go from idea to live app in 3 minutes Klaviyo is a B2C CRM for more than 200,000 brands. It brings marketing, service, agents, and data together in one platform so brands can deliver customer experiences at scale. Klaviyo built a platform on Vercel for shipping internal apps, then opened it to every employee through a citizen develope

<h3>Misc Changes</h3> <ul> <li>[ci] Skip preview tarball upload when the artifact is missing: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5472666204" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/98713" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/98713/hovercard" href="https://github.com/vercel/next.js/pull/98713">#98713</a></li> <li>[cd] Retry preview tarball uploads with backoff

Agents built with TanStack AI can now call OAuth-protected MCP servers through Vercel Connect, with no credentials for you to store or rotate. The new @vercel/connect/tanstack-ai subpath exports connectMCPTransport , which takes a TanStack transport config and attaches a Connect-backed auth provider. The provider is called before every MCP request, so the token is always fresh. If the user has not granted access, createMCPClient fails with a consent challenge before the model runs. Catch it with

<h3>Misc Changes</h3> <ul> <li>Add experimental custom webpack support: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5496194695" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/98862" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/98862/hovercard" href="https://github.com/vercel/next.js/pull/98862">#98862</a></li> <li>Default agentic upgrades to a separate worktree: <a class="issue-link

<h2>What's changed</h2> <ul> <li>Added Claude apps gateway support for newer Claude Desktop keys in <code>desktop</code> policy blocks, including <code>blockReadsOutsideWorkingDirectories</code> and <code>disableBypassPermissionsMode</code></li> <li>Added <code>assume_role</code> on Claude apps gateway Bedrock upstreams: the gateway calls Bedrock as an IAM role it assumes through STS, in another AWS account if needed, optionally one session per developer</li> <li>Added <code>guardrail: {id, vers

You can now create as many Vercel Blob stores as you need. The previous limits of 100 stores on Hobby, 500 on Pro, and 1,000 on Enterprise no longer apply. Blob store creation is now billed alongside other Blob Advanced Operations , including put() , copy() , and list() calls. On Pro that's $5.00 per million. On Hobby it counts toward the 2,000 free operations you get each month. Deleting a store is free. Create a new store whenever you want a hard boundary instead of a pathname convention: Sepa

<img src="https://storage.googleapis.com/gweb-uniblog-publish-prod/images/Google_Beam_hero.max-600x600.format-webp.webp">We’re expanding Google Beam to five new countries, and partnering with Industrious for an extended network.

<h3>Misc Changes</h3> <ul> <li>Keep cached root parameters out of generic fallback shells: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5514477574" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/98950" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/98950/hovercard" href="https://github.com/vercel/next.js/pull/98950">#98950</a></li> <li>Fix missing content in Cached Navigations: <a clas

Gemini 3.8 Flash-Lite TTS and Gemini 3.8 Flash TTS from Google are now available on AI Gateway . Both models take text and generate speech in more than 100 languages. They support long-form narration, control over delivery, and two-speaker dialogue. google/gemini-3.8-flash-lite-tts is suited to high-volume speech generation, with controls for tone, pacing, and line-by-line delivery. google/gemini-3.8-flash-tts adds voice and character design through natural-language prompts, including acting cue

<h3>Misc Changes</h3> <ul> <li>[PPF] Scaffold <code>unstable_ensureStatic</code> segment config: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5327178506" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/98190" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/98190/hovercard" href="https://github.com/vercel/next.js/pull/98190">#98190</a></li> <li>[agent-feedback] Respect CI and telemetry op

<p>This release contains additional security hardening for <code>next/og</code>. For more information, check out <a href="https://nextjs.org/blog/nextjs-security-update-september-22-2026" rel="nofollow">https://nextjs.org/blog/nextjs-security-update-september-22-2026</a></p>

<p>This release contains a security fix for <a href="https://github.com/vercel/next.js/security/advisories/GHSA-vcvr-r3jv-pc5j">GHSA-vcvr-r3jv-pc5j: Remote Code Execution in next/og ImageResponse</a></p>

<h2>What's changed</h2> <ul> <li>Added Claude Opus 5.5 (<code>claude-opus-5-5</code>), now the default Opus model — 1M context, $4/$20 per Mtok with $0.20/Mtok cache reads</li> <li>Added mouse support to more lists in fullscreen mode: the wheel scrolls the <code>/skills</code> list, and a skill's state options in <code>/plugin</code> can be clicked</li> <li>Added <code>CLAUDE_CODE_MAX_MCP_DESCRIPTION_LENGTH</code> to change the 2,048-character cap on MCP tool descriptions and server instructions

<h2>0.0.14 (2026-09-22)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/google-cloud-sdk-v0.0.13...google-cloud-sdk-v0.0.14">google-cloud-sdk-v0.0.13...google-cloud-sdk-v0.0.14</a></p>

<h2>0.7.4 (2026-09-22)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/aws-sdk-v0.7.3...aws-sdk-v0.7.4">aws-sdk-v0.7.3...aws-sdk-v0.7.4</a></p>

<h2>0.4.9 (2026-09-22)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/foundry-sdk-v0.4.8...foundry-sdk-v0.4.9">foundry-sdk-v0.4.8...foundry-sdk-v0.4.9</a></p>

<h2>0.33.8 (2026-09-22)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/bedrock-sdk-v0.33.7...bedrock-sdk-v0.33.8">bedrock-sdk-v0.33.7...bedrock-sdk-v0.33.8</a></p>

<h2>0.19.11 (2026-09-22)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/vertex-sdk-v0.19.10...vertex-sdk-v0.19.11">vertex-sdk-v0.19.10...vertex-sdk-v0.19.11</a></p>

<h2>0.128.0 (2026-09-22)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/sdk-v0.127.0...sdk-v0.128.0">sdk-v0.127.0...sdk-v0.128.0</a></p> <h3>Features</h3> <ul> <li><strong>api:</strong> add support for claude-opus-5-5, inline tool definitions and MCP tool-list pinning (beta) (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/eff27483d20beffef332312ef764a92c80ec5681">eff2748</a>)</li> </ul> <h3>Bug Fixes</h3> <ul> <li><strong>a

<h3>Core Changes</h3> <ul> <li>feat(turbopack): support <code>false</code> values for <code>resolveAlias</code> config: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="4345704837" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/93331" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/93331/hovercard" href="https://github.com/vercel/next.js/pull/93331">#93331</a></li> </ul> <h3>Misc Changes</h

<h2><a href="https://github.com/googleapis/python-genai/compare/v2.24.0...v2.25.0">2.25.0</a> (2026-09-22)</h2> <h3>Features</h3> <ul> <li>Add sample_audio to Voice in GAOS SDK (<a href="https://github.com/googleapis/python-genai/commit/a011633c2db0078019162880764e06cd85a5c4ff">a011633</a>)</li> <li>Add Voices API resource to GAOS SDK (<a href="https://github.com/googleapis/python-genai/commit/c45b94e7a0e0dcd5dcc2590f62f5d223f8c90e99">c45b94e</a>)</li> <li>Expose SpeechMetadata, VoiceConfig.voic

GPT-6 Sol and GPT-6 Luna from OpenAI are now available on AI Gateway . Both models bring GPT-6 improvements in professional work, coding, computer use, factuality, and communication at a lower price than GPT-6 Astra . GPT-6 Sol ( openai/gpt-6-sol ) is suited to complex professional workflows and sustained coding tasks where quality and room to iterate both matter. GPT-6 Luna ( openai/gpt-6-luna ) is the lower-cost option for high-volume agentic workflows, coding, and everyday tasks. Both Sol and

Claude Opus 5.5 from Anthropic is now available on AI Gateway . It is a step-change improvement over Opus 5 , with its biggest gains in agentic coding, long-running agent tasks, and knowledge work. Anthropic cites that Opus 5.5 performs at the level of Fable 5.1, but ~30% faster and ~40% cheaper than Opus 5 per task. Opus 5.5 is also a better collaborator over long runs. It reports back in plain language on what it did, what it found, and what it needs next, making it easier to supervise work th

Drives for Vercel Sandbox are now available in public beta on Hobby, Pro, and Enterprise. A Drive is persistent storage that you mount as a directory in a Vercel Sandbox. It isn’t tied to a single sandbox, so you can reuse the same Drive across runs and different sandbox instances. Use Drives to preserve an agent's workspace or on-disk memory, or to reuse datasets, models, and dependency trees. Create and mount a Drive Create or retrieve a Drive and mount it at a path when starting a sandbox. Re

Vercel Connect now includes a managed connector for Microsoft Teams . Creating one gives your organization a Teams bot that your apps and agents run. People can @mention it in channels or message it directly, and your code receives the message and replies as the bot. As a Vercel Managed Connector , Vercel registers the Entra app and Azure Bot resource in your tenant, so there's no client secret to store. A tenant administrator with an Azure subscription completes setup once. Incoming Teams activ

Deployments now show their billable duration and CPU minutes in the dashboard, vc inspect , and the REST API . Use it to understand how each build contributes to your usage. Billable duration is the build and post-build time combined, rounded up to the next whole minute. CPU minutes are that figure multiplied by the machine's vCPUs. To see the same breakdown from the CLI, update to version 59.23.1 or later with npm i -g vercel@latest , then run vc inspect : Learn more about build machines and CP

MiMo V2.6 Pro , MiMo V2.6 Flash , and MiMo V2.6 Pro UltraSpeed from Xiaomi are now available on AI Gateway . MiMo V2.6 combines coding, reasoning, and tool use with native text, image, audio, and video understanding. Its 1M token context supports long repositories, tool traces, and multi-session agent work, with structured outputs and up to 128K output tokens. Choose a model based on the workload: xiaomi/mimo-v2.6-pro is the larger sparse mixture-of-experts checkpoint, with 1.02T total parameter

You can now call Jev from TypeSafe AI through AI Gateway using an existing TypeSafe client or the HTTP API, in addition to the AI SDK. TypeSafe client: Point an existing TypeSafe client at AI Gateway without changing its evaluation calls. HTTP API: Call Jev directly from any language or framework. AI SDK: Call Jev from a TypeScript application using AI SDK. Jev is a probabilistic decision model for software. State goes in, and typed answers come out with probabilities attached, so there's no gen

Grok 4.7 from SpaceXAI is now available on AI Gateway and 40% off through September 27. The discount applies automatically when you call spacexai/grok-4.7 . Grok 4.7 has a 500K token context window and supports low, medium, high, and xhigh reasoning levels, giving you control over the tradeoff between latency and depth. Use spacexai/grok-4.7 everywhere you call the model. The same ID works with the AI SDK , the OpenAI-compatible Chat Completions API , and coding agents connected to AI Gateway: T

<h3>Misc Changes</h3> <ul> <li>Preserve closed-parameter restrictions in client route prediction: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5501805955" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/98889" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/98889/hovercard" href="https://github.com/vercel/next.js/pull/98889">#98889</a></li> <li>[test] Fix deployment tests that relied on

<h2>What's changed</h2> <ul> <li>Changed auto mode for Claude API and Enterprise users, and on Bedrock, Vertex, Foundry and gateways, to default to the server-side classifier, which does not charge for classifier overhead (<code>CLAUDE_CODE_AUTO_MODE_SERVER=0</code> opts out on Bedrock, Vertex, Foundry and gateways); warns on billed fallback. See <a href="https://code.claude.com/docs/en/auto-mode-classifier-billing" rel="nofollow">https://code.claude.com/docs/en/auto-mode-classifier-billing</a><

<h3>Misc Changes</h3> <ul> <li>Persist resurrected tasks as new tasks: <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5474527830" data-permission-text="Title is private" data-url="https://github.com/vercel/next.js/issues/98729" data-hovercard-type="pull_request" data-hovercard-url="/vercel/next.js/pull/98729/hovercard" href="https://github.com/vercel/next.js/pull/98729">#98729</a></li> <li>Rspack: Remove bincode dependency from rspack by way of turbo-unix-pat

Enterprise teams on Flexible Commitment plans can now use Spend Management , already available on Pro , at no additional cost. You can set a budget at any time in Spend Management settings . Set a budget per billing cycle, and when your team's metered usage approaches or crosses it, Spend Management can send email notifications, trigger a webhook, or pause the production deployments of all projects. The amount governs the metered usage that draws down from your prepaid balance. Setting an amount

<h2>What's changed</h2> <ul> <li>Added AGENTS.md support: in a project with no CLAUDE.md, Claude Code reads AGENTS.md instead; change it under "Project instructions" in <code>/config</code> (not yet on Bedrock, Vertex or Foundry)</li> <li>Added <code>CLAUDE_GATEWAY_PROXY_IS_EGRESS_BOUNDARY=1</code> for Claude apps gateways whose only egress is a forward proxy: every outbound request hands the proxy the hostname instead of resolving it locally</li> <li>Added an optional <code>headers:</code> map

mcp-handler now has experimental support for WebMCP , the proposed web standard for exposing tools to in-browser agents. Add a single script tag to your site, and your existing MCP tools become available there too. Opt tools in by adding them to the experimental_webMcp object: Then load the script from your MCP endpoint with the ?webmcp-script parameter: The script registers those tools with the page and proxies each call back to your MCP server as the signed-in user, so authenticated tools work

v0 now installs private packages from npm and custom registries using credentials stored as shared environment variables on Vercel. This makes it easier for teams to build with their existing design systems, component libraries, and internal packages directly in v0. To get started, add one of the following as a shared environment variable on Vercel, scoped to Development and/or Preview: Use NPM_TOKEN for private packages hosted on registry.npmjs.org . Use NPM_RC to configure custom or multiple r

<h2>0.0.13 (2026-09-18)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/google-cloud-sdk-v0.0.12...google-cloud-sdk-v0.0.13">google-cloud-sdk-v0.0.12...google-cloud-sdk-v0.0.13</a></p> <h3>Chores</h3> <ul> <li><strong>internal:</strong> build the sub-packages with TypeScript 6.0.3 (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/8a69ff594ffeec3f6bd964030086b01ae5354aaf">8a69ff5</a>)</li> <li><strong>internal:</strong> switch

<h2>0.7.3 (2026-09-18)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/aws-sdk-v0.7.2...aws-sdk-v0.7.3">aws-sdk-v0.7.2...aws-sdk-v0.7.3</a></p> <h3>Chores</h3> <ul> <li><strong>internal:</strong> build the sub-packages with TypeScript 6.0.3 (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/8a69ff594ffeec3f6bd964030086b01ae5354aaf">8a69ff5</a>)</li> <li><strong>internal:</strong> switch unit tests from jest to vitest (<a href="

<h2>0.4.8 (2026-09-18)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/foundry-sdk-v0.4.7...foundry-sdk-v0.4.8">foundry-sdk-v0.4.7...foundry-sdk-v0.4.8</a></p> <h3>Chores</h3> <ul> <li><strong>internal:</strong> build the sub-packages with TypeScript 6.0.3 (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/8a69ff594ffeec3f6bd964030086b01ae5354aaf">8a69ff5</a>)</li> <li><strong>internal:</strong> switch unit tests from jest to v

<h2>0.33.7 (2026-09-18)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/bedrock-sdk-v0.33.6...bedrock-sdk-v0.33.7">bedrock-sdk-v0.33.6...bedrock-sdk-v0.33.7</a></p> <h3>Bug Fixes</h3> <ul> <li><strong>bedrock:</strong> raise APIError for mid-stream exception and error frames (<a href="https://github.com/anthropics/anthropic-sdk-typescript/issues/622" data-hovercard-type="issue" data-hovercard-url="/anthropics/anthropic-sdk-typescript/issues/622/hov

<h2>0.19.10 (2026-09-18)</h2> <p>Full Changelog: <a href="https://github.com/anthropics/anthropic-sdk-typescript/compare/vertex-sdk-v0.19.9...vertex-sdk-v0.19.10">vertex-sdk-v0.19.9...vertex-sdk-v0.19.10</a></p> <h3>Chores</h3> <ul> <li><strong>internal:</strong> build the sub-packages with TypeScript 6.0.3 (<a href="https://github.com/anthropics/anthropic-sdk-typescript/commit/8a69ff594ffeec3f6bd964030086b01ae5354aaf">8a69ff5</a>)</li> <li><strong>internal:</strong> switch unit tests from jest