What Vercel shipped in July 2026

Vercel boosts Laguna S 2.1 capacity 10x on AI Gateway

Vercel increased Laguna S 2.1's capacity by 10x on its AI Gateway for both paid and free versions, enabling higher-volume agentic coding and long-running tasks. The update applies to models `poolside/…

Vercel AI Gateway updates DeepSeek V4 Flash with stronger agentic weights

Vercel’s AI Gateway now serves DeepSeek V4 Flash with updated weights by default, boosting its agentic capabilities to a Terminal-Bench score of 82.7 (up 25.8 points). The change is automatic for requ…

Vercel Chat SDK adds reactions and ephemeral messages for Teams

Vercel’s Chat SDK now supports reactions and ephemeral messages in its Microsoft Teams adapter via the same API as other platforms. Bots can add/remove reactions and send targeted, one-time messages v…

Vercel adds structured search and filters for Workflow runs in Observability

Vercel introduced structured search and filtering for Workflow runs in Observability, enabling users to filter by workflow, environment, deployment ID, region, or custom attributes. The feature includ…

Vercel MCP adds support for 2026-07-28 MCP specification

Vercel MCP now supports the 2026-07-28 MCP specification, introducing a stateless request model and updated authorization behavior while maintaining backward compatibility with the 2025 protocol. Both…

Vercel adds dedicated AI Gateway logs page with cost and token tracking

Vercel introduced a dedicated Logs page for AI Gateway, listing all requests with cost, token counts, duration, and routing details. Users can filter by provider, model, or status, export data, and dr…

Vercel AI Gateway adds team and project spend budgets with alerts

Vercel expanded AI Gateway spend controls to support team and project-level budgets alongside API keys, enabling granular spend caps and automatic request blocking when limits are hit. Email alerts at…

Vercel Passport reaches general availability with OIDC integration

Vercel Passport, a deployment protection tool using OIDC identity providers like Okta or Microsoft Entra ID, is now generally available on the Enterprise plan. It injects verified visitor identity tok…

Pricing change detected for Vercel

Pricing updated for Vercel: - Hobby: allowance changed from 100 GB bandwidth/month, 10 GB storage/month, 10K requests/month, 1 project to 100 GB bandwidth/month, 10 GB storage/month, 1K ISR reads/mont…

Shopify and Vercel partner to rebuild Hydrogen for faster storefronts

Shopify and Vercel are rebuilding Hydrogen, Shopify's headless storefront framework, to be open source and runtime agnostic. The new version supports frameworks like Next.js and Nuxt, reducing custom …

Vercel AI Gateway cuts GPT-5.6 Luna pricing by 80% and Terra by 20%

Vercel reduced pricing for GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20% on its AI Gateway, aligning with upstream rates. GPT-5.6 Sol’s fast mode speed doubled to 2.5x while maintaining the same price.…

Vercel to pass through Server-Timing headers to clients starting August 10, 2026

Vercel will begin passing through Server-Timing response headers to clients on August 10, 2026, enabling developers to expose backend metrics like database query time and cache hits in browser tools. …

Vercel launches project-scoped tokens for API authentication

Vercel introduced project-scoped tokens that restrict API access to a single project, preventing cross-project or user/team-level resource access. Tokens are created via the Account Tokens page with p…

Vercel enables Flexible Commitment for Marketplace purchases with Neon, Supabase, Redis

Vercel introduced Enterprise Flexible Commitment support for eligible Marketplace resources, allowing customers to apply committed spend toward purchases from partners like Neon, Supabase, and Redis. …

Vercel mcp-handler 2.0 adds support for latest MCP spec and TypeScript SDK v2

Vercel released mcp-handler@2.0.0 with support for the 2026-07-28 Model Context Protocol specification and MCP TypeScript SDK v2. The update enables stateless protocol handling, removes Redis dependen…

Vercel deployments up to 7 seconds faster with platform optimizations

Vercel reduced deployment times by up to 7 seconds end-to-end by removing platform overhead and optimizing CLI interactions. Changes include faster build-process shutdown, concurrent metadata processi…

Vercel adds OIDC support for Turborepo and Remote Cache

Vercel now supports OpenID Connect (OIDC) tokens in Turborepo and Remote Cache, replacing long-lived Personal Access Tokens (PATs) with short-lived, team-scoped tokens for CI/CD workflows. This enhanc…

Vercel adds Inkling Small from Thinking Machines to AI Gateway

Vercel’s AI Gateway now supports Inkling Small, a compact model from Thinking Machines offering performance comparable to larger models at 25% the size with native audio/image reasoning. It enables co…

Vercel Sandbox SDK adds multi-user isolation for agents

Vercel’s @vercel/sandbox SDK now supports multiple Linux users and groups, enabling isolated agents to run side by side in a single Sandbox. Each agent operates as a distinct user with private home di…

Vercel adds MiniMax H3 video model to AI Gateway

Vercel integrated MiniMax H3, a 2K video generation model, into its AI Gateway, supporting text-to-video, image-to-video, keyframe transitions, and multimodal reference-based generation. Outputs are 5…

Vercel rebrands Edge Config to Global Config with upgraded limits and new SDK

Vercel renamed Edge Config to Global Config to better reflect its globally replicated 1ms-read data store for runtime configurations like feature flags. The new `@vercel/global-config` SDK replaces `@…

Vercel AI Gateway adds unified fast mode support for all models

Vercel’s AI Gateway now offers a unified fast mode abstraction in beta, allowing developers to request faster serving tiers across all supported models with a single `speed: 'fast'` parameter. Fast va…

Vercel adds Grok Voice Think Fast 2.0 to AI Gateway

Vercel now offers Grok Voice Think Fast 2.0 from xAI on its AI Gateway, a speech-to-speech model that processes audio in and out with improved reasoning, transcription accuracy, and parallel thinking …

Vercel enables CLI-based discovery and installation of eve integrations

Vercel introduced CLI commands (`eve add`, `eve registry list/search/view`) to discover, inspect, and install eve agent integrations directly from the terminal. Users can add third-party sources via n…

Vercel adds paid custom environments for Pro and Enterprise teams

Vercel now lets Pro and Enterprise teams purchase additional custom environments (packs of 5) for $50/month per pack via dashboard, API, or CLI. Pro plans include 1 environment per project (up to 16 t…

Vercel Connect adds support for Custom Environments in beta

Vercel Connect now supports linking connectors to Custom Environments, enabling deployments in those environments to request provider tokens and receive forwarded webhooks. Custom Environments appear …

Vercel Sandbox adds forking capability for multi-tenant and template branching

Vercel Sandbox now supports forking via `Sandbox.fork()`, enabling users to create independent copies of sandboxes with inherited configurations and environment variables. Forks start from the latest …

Vercel enables Claude Managed Agents with Chat SDK for multi-platform chat interfaces

Vercel introduced an integration allowing Claude Managed Agents to operate through Chat SDK, providing token-by-token streaming, live activity feeds, and no server-side state management. The setup sup…

Sandstone grows 40x in 147 days using Vercel’s AI SDK and Secure Compute

Sandstone, a legal workflow automation platform, achieved 40x revenue growth in 147 days by leveraging Vercel’s infrastructure, including AI SDK for agentic workflows, Secure Compute for private data …

Vercel launches DeepsecBench to benchmark AI models for vulnerability detection

Vercel introduced DeepsecBench, a benchmark evaluating AI models' ability to detect cybersecurity vulnerabilities in code, with performance metrics like recall, precision, cost, and time. The tool aim…

Vercel AI Gateway adds regional inference for US and EU data residency

Vercel’s AI Gateway now supports regional inference with pinned regions (US or EU) via the `inferenceRegion` parameter, ensuring data residency compliance. Requests fail if no provider supports the se…

Vercel mitigates Nuxt RCE and other vulnerabilities with platform WAF and patches

Vercel deployed platform-wide WAF mitigations for a high-severity Nuxt server-side RCE vulnerability before public disclosure, protecting deployed apps automatically. Nuxt released patches (4.5.1, 3.2…

Vercel’s eve adds Slack session controls and event hooks

Vercel’s eve AI agents on Slack now support autonomous thread replies without mentions, session cancellation, and conversation resets via new `ctx.cancel()` and `ctx.reset()` helpers. The update also …

Vercel AI Gateway adds WebSocket support for OpenAI Responses API

Vercel’s AI Gateway now supports WebSocket mode for the OpenAI Responses API, enabling persistent connections to reduce context re-sending and improve performance. OpenAI reports up to 40% faster exec…

Vercel adds Moonshot AI's Kimi K3 models to AI Gateway with US-based providers and ZDR

Vercel integrated Moonshot AI’s Kimi K3 and Kimi K3 Fast models into its AI Gateway, now available via US-based providers like Baseten and Fireworks. Both models support Zero Data Retention (ZDR), and…

Vercel adds Anthropic's Claude Opus 5 to AI Gateway with coding agent support

Vercel integrated Anthropic's Claude Opus 5 into its AI Gateway, enhancing long-horizon coding tasks, multi-file refactors, and visual analysis with stronger tool coordination. The model supports vari…

Vercel extends workflow step durations to 30 minutes for Pro and Enterprise plans

Vercel increased the maximum duration for workflow steps on Pro and Enterprise plans from 800 to 1800 seconds (30 minutes), requiring an opt-in via environment variable and Fluid compute. Hobby plans …

Vercel WAF for Blob enters beta with edge-based protection

Vercel extended its Web Application Firewall (WAF) to protect Vercel Blob stores in beta, allowing edge-based rules (deny, challenge, rate limit, redirect) to be applied to blob traffic without code c…

Vercel adds Ling 3.0 Flash to AI Gateway with free trial

Vercel introduced Ling 3.0 Flash, a 124B-parameter Mixture-of-Experts model with 256K context window, now available on Vercel’s AI Gateway. The model is free to use until August 3, 2026, and targets h…

Vercel MCP now supports direct code deployment from AI assistants

Vercel's MCP server now enables AI assistants to deploy code directly to new or existing projects, automatically detecting frameworks, installing dependencies, and providing a shareable URL without le…

Vercel adds GitHub tools as an installable eve extension

Vercel introduced a new GitHub tools extension for its eve agent platform, allowing users to integrate 42 GitHub tools via a simple install and configuration process. The extension leverages Vercel Co…

Vercel CLI adds feature flag version history inspection

Vercel introduced a new `vercel flags versions` CLI command to inspect the full revision history of feature flags, including author, message, timestamp, and environment changes. Users can filter, pagi…

Vercel adds Sandbox management from the dashboard

Vercel introduced a new dashboard feature allowing users to connect to running Sandboxes directly from the browser, enabling command execution, file management, and port inspection without leaving the…

Vercel Flags adds live evaluation metrics with granular filtering

Vercel Flags now provides a live evaluation view on each flag's detail page, showing per-minute metrics over time with version markers. Users can filter and group evaluations by variant, environment, …

Pricing updated for Vercel: - Hobby: allowance changed from 1 project to 100 GB bandwidth/month, 10 GB storage/month, 10K requests/month, 1 project - Pro: promotion changed — $20 included credit to us

Vercel enables installable extensions for eve agents

Vercel introduced installable extensions for eve agents, allowing tools, connections, skills, and hooks to be packaged and shared via registries. Developers can scaffold, build, and publish extensions…

Vercel AI Gateway adds streaming transcription for real-time audio processing

Vercel’s AI Gateway now supports streaming transcription, enabling real-time transcript updates as audio is captured. This reduces latency for live captioning and voice input use cases. The feature is…

Vercel AI Gateway adds service tiers for latency, throughput, and cost optimization

Vercel introduced service tiering for AI Gateway, allowing users to optimize requests for latency, throughput, or cost. Three tiers are available: default, priority (faster but ~1.8-2x cost), and flex…

Vercel MCP adds in-app purchasing for Pro plans, v0 credits, AI Gateway, SIEM, and domains

Vercel MCP now enables AI assistants to directly purchase Vercel products, including Pro plan upgrades, v0 credits, AI Gateway usage, SIEM add-ons, and domains. Transactions require billing access and…

Vercel adds Laguna S 2.1 to AI Gateway with 1M token context window

Vercel’s AI Gateway now supports Laguna S 2.1, an open-weight Mixture-of-Experts model from Poolside, with a free 256K context version and a paid 1M context version. The model excels in agentic coding…

Vercel Connect expands to 90+ preset connectors including Shopify, Okta, and Workday

Vercel Connect now offers 90+ preset connectors for services like Shopify, Okta, Workday, Jira, and Sanity, simplifying setup with pre-configured auth and discovery details. Users can add connectors v…

Vercel adds Google Gemini 3.6 Flash and 3.5 Flash-Lite to AI Gateway

Vercel integrated Google’s Gemini 3.6 Flash and 3.5 Flash-Lite models into its AI Gateway, improving coding, agentic tasks, and web development efficiency while reducing token usage. The models are ac…

Searchable ships features 5x faster on Vercel’s AI tools

Searchable, an AI search visibility platform, reduced feature development time from hours to 30 minutes by leveraging Vercel’s AI SDK and AI Gateway. This enabled 5x faster development velocity and ze…

Vercel Python functions now ship with precompiled bytecode

Vercel’s Python function bundles now include precompiled bytecode, reducing median cold starts from 2.8s to 1.3s. The change skips runtime compilation by bundling `.pyc` files, with no code changes re…

Pricing updated for Vercel: - Hobby: allowance changed from 100 GB bandwidth/month, 10 GB storage/month, 10K requests/month, 10K ISR reads/month, 1K ISR writes/month, 1 project, 1 environment, 1 concu

Vercel adds one-click Remote Cache purge for build and CI artifacts

Vercel introduced a feature allowing Team Owners to clear their team's Remote Cache of all build and CI artifacts in one click, addressing concerns about poisoned artifacts. The option is available in…

Vercel Workflows now pins run state to a single home region

Vercel Workflows now defaults to keeping each run’s state, queue dispatch, and output streams in the region where the run starts, with an option to pin runs to a specific region. This ensures continui…

Vercel optimizes CDN caching for immutable static assets with 60%+ TTFB improvement

Vercel introduced zero-config optimization for immutable static assets, reusing files across deployments to reduce CDN requests by 17% and bytes transferred by 24%. Deployments are up to 30% faster, a…

Vercel offers 35% off GLM 5.2 via Novita on AI Gateway until July 24

Vercel announced a 35% discount on GLM 5.2 through its AI Gateway when routed via Novita, valid until July 24, 2026. After the date, the model will revert to standard provider rates without markup.

Vercel Plugin now available in Kimi Code CLI

Vercel announced its official plugin is now accessible within the Kimi Code CLI, enabling AI-driven access to Vercel’s platform knowledge including Next.js, AI SDK, and Vercel Functions. Users can ins…

Vercel Sandbox removes data download charges from billing

Vercel Sandbox no longer bills for data downloaded from the internet, including package installations, Git clones, or external dataset pulls. Outbound traffic and exposed port traffic remain billable,…

Vercel runtime logs now show cache reasons for CDN debugging

Vercel added cache reasons to runtime logs to explain why requests miss the CDN cache, such as time-based or tag-based revalidation. This applies to ISR, Partial Prerendering, and responses with Cache…

Vercel Chat SDK adds native Slack agent support with Agent Messages

Vercel's Chat SDK now supports Slack's Agent Messages experience, including Agent badges, Messages tab integration, suggested prompts, rotating status messages, and native feedback buttons. The update…

Searchable ships features 5x faster using Vercel’s AI SDK and AI Gateway

Searchable, an AI search visibility platform, reduced feature development time from hours to 30 minutes by leveraging Vercel’s AI SDK and AI Gateway. This enabled 5x faster development velocity and pr…

How Vercel helps mmm.page manage over 30,000 sites

mmm.page migrated from Amazon S3 to Vercel to handle 30,000+ custom domains, citing SSR for SEO and scalability, easier custom domain setup, and improved Core Web Vitals. The hybrid architecture uses …

Vercel adds Moonshot AI's Kimi K3 to AI Gateway with 1M-token context

Vercel integrated Moonshot AI’s open-source Kimi K3 model into its AI Gateway, offering a 1M-token context window with native visual understanding for text, images, and video. The model targets long-h…

Vercel adds write utilization metric to ISR Observability

Vercel introduced a write utilization metric in its ISR Observability tool for Observability Plus subscribers, helping identify routes with high ISR write frequency but low request volume. This enable…

How Makeswift improved CI speed by 65% with Turborepo

Makeswift, a Next.js site builder, reduced CI pipeline time by 65% and cut commit-to-deploy latency from 20 minutes to under one minute by adopting Vercel’s Turborepo. The switch eliminated manual tas…

Speechify case study: Scaling 500K dynamic pages to 60M users on Vercel

Speechify migrated to Vercel’s platform, cutting costs 50% while serving 500K+ dynamic pages in 40+ languages to 60M users. They achieved 99.99% uptime, instant rollbacks, and continuous deployments, …

Vercel Workflows trace viewer gains minimap for navigation

Vercel added a minimap to its Workflows trace viewer, enabling users to navigate entire runs via drag-to-pan, resize-to-zoom, or click-to-jump. The feature is available in the Vercel UI and locally vi…

Vercel Chat SDK adds Discord Components V2 support with new UI features

Vercel’s Chat SDK now supports Discord’s Components V2 layout system, enabling flexible message composition with native containers, sections, media galleries, and interactive buttons. The update also …

Vercel Connect now integrates with GitHub Tools for scoped, short-lived GitHub tokens

Vercel added first-class support for Vercel Connect in GitHub Tools, replacing long-lived personal access tokens with short-lived, scoped tokens minted at runtime. Developers can attach a GitHub conne…

Vercel AI Gateway adds Inkling from Thinking Machines to its model catalog

Vercel’s AI Gateway now supports Inkling, a generalist AI model from Thinking Machines, enabling developers to access it via a unified API with usage tracking, cost management, and failover features. …

Vercel CLI v55 speeds up project linking with team-first resolution

Vercel CLI v55 introduces team-first project resolution, reducing link discovery time by prioritizing team context and local Git matches before broader searches. Non-interactive commands now require e…

Speechify scales to 60M users on Vercel with 500K dynamic pages

Speechify migrated to Vercel’s platform, cutting costs 50% while serving 500K+ dynamic pages in 40+ languages to 60M users. They achieved 99.99% uptime, instant rollbacks, and continuous deployments, …

Vercel Workflows: Durable apps and AI agents with managed infrastructure

Vercel Workflows is a fully managed platform for building durable, reliable, and observable applications and AI agents using JavaScript, TypeScript, and Python. It enables code to pause, resume, and m…

Vercel expands Spend Management with webhooks, SMS alerts, and auto-pause

Vercel introduced expanded Spend Management capabilities for Pro plan teams, allowing automatic notifications at 50%, 75%, and 100% of a set spend threshold, SMS alerts at 100%, and webhook triggers f…

How a global agency built a web innovation engine in two months

Vercel highlights a case study featuring a global agency that developed a web innovation engine in a two-month timeframe using their platform.

How Supabase elevated their developer experience with Turborepo

Vercel highlights a case study featuring Supabase, demonstrating how they improved developer experience using Turborepo, Vercel's build system.

How Fern delivers 6M+ monthly views and 80% faster docs with Vercel

Fern, a multi-tenant API documentation platform, migrated to Vercel’s infrastructure and Next.js App Router, cutting page load times by 50-80% and reducing deploy times from 30 minutes to five. The sh…

WordPress monolith to Vercel: How Personio elevated site performance and efficiency

Personio, Europe’s leading HR solution provider, migrated from a monolithic WordPress stack to a composable architecture using Vercel, Next.js, and Contentful. The shift improved Core Web Vitals by ov…

Vercel Chat SDK adds X adapter support for bot development

Vercel's Chat SDK now includes an X (Twitter) adapter, enabling bots to respond to public mentions and direct messages via the X API v2. The adapter automates CRC verification, webhook signatures, and…

Vercel Marketplace adds AgentMail for AI agent email management

Vercel integrated AgentMail into its marketplace, allowing AI agents to send, receive, and respond to emails via a real inbox with built-in deliverability. Users can provision inboxes and domains via …

Vercel to deprecate Node.js 20 for Functions and Builds by October 1, 2026

Vercel will deprecate Node.js 20 for Builds and Functions on October 1, 2026, following its upstream end-of-life on April 30, 2026. Existing deployments remain unaffected, but new deployments will fai…

Endform joins Vercel Marketplace for Playwright test execution

Vercel added Endform to its Marketplace, enabling parallel Playwright test execution with isolated machines and pass/fail checks on deployments. Users can install via Vercel CLI and pay per test runti…

Vercel Blob adds consistent reads for private storage with useCache: false

Vercel Blob now supports consistent reads on private storage by passing `useCache: false` to `get()` or `presignUrl()`, ensuring reads reflect the latest write immediately. Overwritten blobs may retur…

Vercel Plugin now available in VS Code and GitHub Copilot CLI

Vercel launched a plugin for GitHub Copilot that embeds Vercel platform knowledge, including Next.js, AI SDK, and Vercel Functions. The plugin is installable via VS Code Extensions or the GitHub Copil…

How Superset built the IDE for AI agents on Vercel

Superset, an IDE for multi-agent AI development, runs entirely on Vercel’s platform, handling 1,000–1,400 deployments weekly and 600 preview deployments daily with 30-second average build times. The t…

How Okara runs CMO agents for 120,000 companies on Vercel

Okara, an AI CMO platform serving 120,000+ businesses, adopted Vercel AI Gateway to unify eight AI model providers under one API, eliminating custom SDK integrations and enabling same-day model availa…

Vercel adds open data and shareable charts to AI Gateway leaderboards

Vercel enhanced its AI Gateway leaderboards with open data access and shareable chart exports. Users can now download CSV files or query anonymized data programmatically via a new API endpoint, enabli…

Vercel launches Deployment Policies to restrict deployment sources by environment

Vercel introduced Deployment Policies, allowing teams to restrict which sources (mechanisms, organizations, repositories) can create deployments per environment at both team and project levels. This e…

Vercel reduces Web Analytics and Speed Insights costs by ~10%

Vercel made Web Analytics and Speed Insights roughly 10% more cost-efficient for Pro and Enterprise teams by improving infrastructure and event-processing efficiency. The changes apply automatically w…

Vercel Agent Runs adds subagent activity tracking for Eve projects

Vercel introduced a Subagents tab in Agent Runs to display subagent activity for Eve projects, showing prompts, durations, failures, and a shared timeline. Each subagent run can be inspected with full…

Vercel AI Gateway July 2026: Open-weight models surge to 29% of token volume, price per token flattens

Vercel's AI Gateway reported a 29% month-over-month growth in token volume and 27% in spend for June 2026, with open-weight models accounting for 29% of tokens (up from 11% in April) on just 4% of spe…

Vercel adds CLI support for managing Flags targeting rules

Vercel introduced a new CLI command (`vercel flags rules`) to manage targeting rules for Vercel Flags directly from the terminal. Users can add, move, or inspect rules, with full parity to the dashboa…

Delivering AI analysis faster with the Vercel workflow

AI analytics firm Viable reduced feedback bottlenecks by 80% by adopting Vercel’s Preview Deployment comments feature, eliminating lost context in Slack/email/Zoom. Their Next.js-based product, deploy…

How Vercel and Next.js keep Rippling on their rising path to success

Rippling, an HR platform scaling from $13M to $100M in revenue, replaced a fullstack WordPress solution with Next.js and Vercel to handle rapid iteration and performance needs. Over 90% of site change…

Using the AI SDK to build Sitecore Stream's AI-powered brand aware assistant

Sitecore launched Sitecore Stream, an AI-powered brand assistant for marketers, built using Vercel’s AI SDK with Python backend, Next.js frontend, and real-time streaming. The integration enabled rapi…

Never drop the illusion: How Frame.io builds fluid user experiences

Frame.io, an Adobe company, adopted Vercel’s platform and Next.js App Router to achieve sub-100ms UI response times and 60fps animations for video professionals. Their 'Fluid UI' principles—Instant, S…

Elkjøp's Digital Transformation: Powering Retail Innovation with Next.js and Vercel

Elkjøp, a Nordic consumer electronics retailer, migrated from Angular to Next.js and Vercel to address slow page loads, poor SEO, and inefficient workflows. The shift improved Core Web Vitals by 50-93…

Konabos empowers an industry giant to deploy 50% faster with a composable stack

Konabos helped American Bath Group replace a monolithic setup with a composable stack (Vercel, Next.js, Kontent.AI), reducing deployment time by 50%, cutting build times from 30 hours to 3 minutes, an…

Tray.ai cut build times from a day to minutes with Vercel

Tray.ai migrated its marketing site from a legacy Next.js setup to Vercel, reducing build times from 24 hours to two minutes. The new architecture consolidated 500K+ pages into a single Turborepo mono…

Vercel adds Seedream 5.0 Pro to AI Gateway

Vercel now offers Seedream 5.0 Pro, a text-to-image and editing model, on its AI Gateway platform. The model excels at rendering text without errors and creating dense infographics, and is accessible …

Developing at the speed of sound: How Sonos amplified their DevEx

Sonos migrated to a headless Next.js architecture on Vercel, reducing build times by 75% and improving mobile Lighthouse scores to 90. The switch enabled faster deployments, better observability, and …

Achieving feature rollouts with ultra-low latency and zero impact to conversion

Vercel highlights its integration with Hypertune, enabling seamless feature flag and A/B testing management for Next.js apps deployed on Vercel. The solution leverages Edge Config for ultra-low latenc…

Integrating Vercel and Sitecore for 2x faster development times and 111% higher conversions

Avanade adopted Vercel and Sitecore XM Cloud to replace monolithic systems, integrating with Microsoft’s Azure DevOps for a composable architecture. The stack improved performance, reduced carbon emis…

Pricing updated for Vercel: - Hobby: allowance changed from 100 GB bandwidth / month, 10 GB storage / month, 10K requests / month, 1 project to 100 GB bandwidth/month, 10 GB storage/month, 10K request

Vercel Logs now support Tree and Waterfall views for traces

Vercel added Tree and Waterfall views to its Traces feature in Vercel Logs, enabling developers to visualize span hierarchies, critical paths, and timing details directly within log entries. Tree view…

Retailer sees $10M increase in sales on Vercel

A top global sportswear retailer migrated from Salesforce Commerce Cloud to Vercel, addressing performance bottlenecks and scalability issues ahead of Black Friday. The move resulted in a $10M revenue…

Hashnode runs the fastest blogs on the web with Vercel

Hashnode, a Next.js-based developer blogging platform, migrated from self-managed infrastructure to Vercel to handle custom domains at scale and improve performance. The move leverages Vercel’s Edge N…

Pricing updated for Vercel: - Pro: promotion changed — Includes $20 monthly credit for usage - Pro: allowance changed from 100 GB bandwidth / month, 100 GB storage / month, 1M requests / month, 12 env

Creating a robust platform for documentation with Next.js and Vercel

Teleport rebuilt its documentation and website using Next.js and Vercel to replace MkDocs and Hugo, enabling dynamic content, automated GitHub deployments, and real-time CMS updates. The migration imp…

How FLORA shipped a creative agent on Vercel's AI stack

FLORA introduced FAUNA, an AI agent that automates creative workflows by generating variations of visual ideas from simple prompts, replacing their node-based canvas. Built on Vercel's AI Stack, FAUNA…

Vercel adds zero-config support for Lovable apps via TanStack Start

Vercel now supports deploying Lovable applications with zero configuration by automatically detecting the TanStack Start framework and deploying via Nitro. Changes in Lovable sync to GitHub and trigge…

Wall Street Journal: China Weighs Limits on the AI Models American Companies Love

9 Jul 2026 Perspectives ↗ Wall Street Journal: China Weighs Limits on the AI Models American Companies Love Across Silicon Valley, models made by Chinese companies such as DeepSeek and Moonshot AI are…

Vercel adds Meta’s Muse Spark 1.1 to AI Gateway

Vercel integrated Meta’s Muse Spark 1.1 into its AI Gateway, enabling multimodal agentic tasks with a 1M token context window. The model supports parallel tool calls, structured output, and built-in s…

Vercel AI Gateway adds OpenAI GPT 5.6 models Sol, Terra, and Luna

Vercel’s AI Gateway now supports OpenAI’s GPT 5.6 models—Sol, Terra, and Luna—offering improved agentic performance, token efficiency, and cost savings. Users can route traffic to these models via the…

Vercel build logs now redact sensitive environment variable values

Vercel now automatically redacts values of sensitive environment variables (32+ characters) in build logs, replacing them with [REDACTED]. The change applies to both user-defined sensitive variables a…

How Waldium made a blog platform work for humans and AI alike

Waldium, a YC-backed startup, built an agentic CMS with MCP endpoints for AI agents to query blogs directly. The platform serves 500+ customer blogs from a single Next.js deployment on Vercel, reducin…

Revolutionizing video editing on the web with Next.js and Vercel

Ozone, a web-based video editing startup, migrated from Heroku to Vercel, cutting deployment times from 15 minutes to seconds and accelerating development by 5x. Vercel’s branching and AI integration …

Tekla's ecommerce evolution: harnessing flexibility with Vercel and Medusa

Tekla, a Copenhagen-based bedding brand, migrated from Gatsby Cloud and WooCommerce to Vercel and Medusa to handle high traffic, improve performance, and enable personalized experiences. The new stack…

How Dub grew to 3,000 active domains with Vercel’s multi-tenant SaaS toolkit

Dub, an open-source link management platform, grew to 3,000 active domains and 5 million monthly redirects by leveraging Vercel’s Domains API and managed infrastructure. Vercel simplified domain manag…

SERHANT.'s playbook for rapid AI iteration

SERHANT. built its AI-powered real estate platform S.MPLE on Vercel’s Next.js stack, expanding to a React Native iOS app without backend changes. Using Vercel’s AI SDK and AI Gateway, the team orchest…

How Notion Workers run untrusted code at scale with Vercel Sandbox

Notion launched Notion Workers, enabling developers to extend Custom Agents with custom code for data syncing, automations, and API calls, all running on Vercel Sandbox. Vercel Sandbox provides hard i…

Vercel Chat SDK adds Photon support for iMessage bots

Vercel’s Chat SDK now includes an official Photon adapter enabling bots to send and receive iMessages, including group chats, media sharing, and native tapbacks. The adapter runs against Spectrum Clou…

Vercel launches Agent for autonomous production investigation and fixes

Vercel introduced Vercel Agent, an AI-powered assistant that autonomously investigates production issues, proposes fixes, and executes approved actions like rollbacks or config changes. It operates un…

Vercel CLI gains project settings updates without dashboard access

Vercel added CLI commands to modify project framework presets and build settings directly, eliminating the need to use the dashboard. Users can now run `vercel project update` with flags like `--frame…

Vercel adds Microfrontends config check to block broken production deployments

Vercel introduced a new deployment check that blocks default app deployments missing the required `microfrontends.json` file, preventing broken production routing. The check is active by default for e…

Vercel Flags SDK evaluates flags 10x faster with bulk evaluation

Vercel’s Flags SDK and Vercel Flags now evaluate feature flags in bulk up to 10x faster by reducing microtask queue overhead and promise creation. Developers can use `evaluate([flagA, flagB])` or name…

Vercel adds Grok 4.5 to AI Gateway for unified model access

Vercel integrated SpaceXAI’s Grok 4.5 into its AI Gateway, enabling developers to access the model via a unified API with configurable reasoning levels. AI Gateway provides usage tracking, cost manage…

Vercel Chat SDK integrates with Vercel Connect for credential management

Vercel introduced support for Vercel Connect in its Chat SDK, enabling credential management for bots without manual token storage or rotation. The integration includes adapter helpers for Slack, GitH…

Vercel Chat SDK adds Dial support for SMS, MMS, and voice transcripts

Vercel’s Chat SDK now includes an official adapter for Dial, enabling bots to send and receive SMS, MMS, and iMessage via real phone numbers with bidirectional media and voice-call transcript support.…

Vercel's eve launches Chat SDK channel for multi-platform messaging integrations

Vercel's eve introduced a new Chat SDK channel enabling eve agents to connect with Facebook Messenger, WhatsApp, Resend, Liveblocks, and other platforms via adapters. Developers can now use standard C…

How Coxwave delivers GenAI value faster with Vercel

Coxwave, a GenAI consulting and analytics platform, reduced deployment times by 85% and increased production deploys from weekly to twice daily by migrating to Vercel and Next.js. Preview Deployments …

Pricing updated for Vercel: - Hobby: now includes 100 GB bandwidth / month, 10 GB storage / month, 10K requests / month, 1 project - Pro: new promotion — $20 included credit to use across resources -

BNP Paribas Open: Serving up scores and experiences in real time with Work & Co and Vercel

The BNP Paribas Open, a major tennis tournament, replaced outdated CMS platforms with Vercel’s Next.js and edge network to deliver real-time scores, personalized content, and seamless web/mobile exper…

Extra Space Storage's build times became 17x faster with Vercel

Extra Space Storage, the largest U.S. self-storage company, migrated its flagship website and kiosk platform to Vercel, reducing build times by 95% and improving site reliability. The shift to Vercel’…

How Zapier scales product partnerships with v0

Zapier leverages Vercel's v0 AI tool to generate high-fidelity, interactive prototypes for its Powered by Zapier integration platform in under two hours, replacing weeks of manual design work. This ac…

Vercel Sandbox gains granular observability with CPU, memory, and data transfer metrics

Vercel added detailed resource metrics to Sandbox observability, including active CPU, provisioned memory, data transfer, and running sessions. Metrics align with billing and can be queried via CLI or…

Vercel adds GitHub Tools integration for eve agent framework

Vercel introduced a new `@github-tools/sdk/eve` subpath enabling eve agents to register GitHub tools with minimal code, including presets like 'maintainer' for full GitHub agent functionality. Write t…

Vercel acquires Better Auth to accelerate open source auth and agent identity

Vercel acquired Better Auth, the company behind the open-source TypeScript authentication library with 4.7M+ weekly npm downloads and 850+ contributors. The Better Auth team, including founder Bereket…

Vercel adds project-level Activity Log in Project Settings

Vercel introduced a project-specific Activity Log within Project Settings, aggregating relevant user events from the global Team Settings log. Users can now review project-level activity without navig…

CNBC: Chinese AI models are gaining ground with U.S. companies as OpenAI, Anthropic costs surge

7 Jul 2026 Perspectives ↗ CNBC: Chinese AI models are gaining ground with U.S. companies as OpenAI, Anthropic costs surge Harpreet Arora, head of agentic infrastructure, shares insights from Vercel&#x…

Vercel CEO Guillermo Rauch on the fight to split off models from agents

Vercel CEO Guillermo Rauch revealed two new frameworks—Eve for agent instruction management and Vercel Sandbox for secure data access control—to address production challenges with AI agents. These too…

How Avalara turns pipe dreams into patent-pending with v0

Avalara’s Chief Strategy and Product Officer Jayme Fishman used Vercel’s v0 to turn plain-language ideas into working prototypes in days, enabling two patent-pending products in 60 days. The tool repl…

How Code and Theory cut time-to-prototype 75% with v0

Digital agency Code and Theory reduced time-to-prototype by 75% and deployment timelines by over 50% by replacing wireframes with Vercel's v0 prompt-to-code workflows, enabling real-time prototype cre…

How Fern runs multi-tenant docs for Webflow and ElevenLabs on Vercel

Fern, a platform for developer documentation and SDKs, migrated 65% of its platform from Next.js Pages Router to App Router in seven days, reducing page load times by up to 80% and achieving 3x faster…

How OpenEvidence built a healthcare AI that physicians actually trust

Vercel highlights a case study of OpenEvidence, a healthcare AI platform, demonstrating its reliability and trustworthiness for physicians. The content showcases Vercel's role in enabling the solution…

How Ramp kept 100% uptime through 100x traffic surges on Vercel

Ramp rebuilt its marketing site on Next.js and Vercel, migrating high-traffic pages with A/B testing against Core Web Vitals. The site maintained 100% uptime during 100x traffic surges and implemented…

How Stripe built a game-changing app in a single flight with v0

Stripe’s Business Value Consulting team used Vercel’s v0 to prototype a customer-facing value calculator in a single flight, replacing months of spreadsheet-based workflows. The tool now sees 288% mor…

How Stably ships AI testing agents in hours, not weeks

Stably, a 6-person AI testing startup, reduced deployment time from weeks to hours using Vercel’s platform, enabling rapid feature shipping and enterprise growth. Their participation in Vercel’s AI Ac…

How Sensay went from zero to product in six weeks

Vercel highlights a customer case study featuring Sensay, which developed a product from scratch in six weeks using Vercel's platform. The story emphasizes rapid development and deployment.

Managing 275 thousand pages and 8 million assets at top speed with ISR

Mecum Auction rebuilt its digital platform using Vercel’s Frontend Cloud with Incremental Static Regeneration (ISR) to handle 275,000 listing pages and 8 million assets, enabling on-demand page genera…

How the Weather Company serves real-time forecasts to 350 million daily active users on Vercel

The Weather Company rebuilt its web serving stack and CMS on Vercel and v0, reducing time from design to live page from days to hours and achieving an 80% increase in velocity. The platform now serves…

Vercel Sandbox adds FUSE-based filesystem support for remote storage

Vercel Sandbox now supports FUSE-based filesystems, enabling users to mount remote storage like S3 buckets or network filesystems directly within a running Sandbox. This allows streaming large dataset…

Vercel adds MCP tools and CLI commands for eve Agent Runs observability

Vercel introduced new MCP tools and CLI commands to inspect eve Agent Runs, enabling teams to list projects with activity, retrieve recent runs, and access full traces including reasoning, tool calls,…

Vercel CLI adds segment management for Vercel Flags

Vercel introduced a new `vercel flags segments` CLI command to manage feature flag segments directly from the terminal. Users can incrementally edit segments using `--add`, `--remove`, or replace them…

Vercel adds Agent Runs observability to MCP and CLI for eve framework

Vercel introduced Agent Runs observability in its MCP and CLI for the open-source eve agent framework, enabling teams to inspect, list, and trace agent runs with full metadata, reasoning, and token us…

Pricing updated for Vercel: - Plan features or structure changed

Vercel AI Gateway adds routing rules for model failover and access control

Vercel AI Gateway now supports routing rules, enabling teams to rewrite or deny model requests at the gateway level without code changes. Rewrite rules reroute traffic from unavailable or retired mode…

Mercedes-AMG PETRONAS F1 Team and Vercel Announce Multi-Year Strategic Partnership

Mercedes-AMG PETRONAS F1 Team and Vercel formed a multi-year strategic partnership to leverage Vercel’s agentic infrastructure for building high-performance digital experiences. The collaboration aims…

Pricing updated for Vercel: - Hobby: now includes 100 GB bandwidth / month - Pro: allowance changed from 1 seat included to 3 seats included

Vercel restores access to Claude Fable 5 on AI Gateway with updated safety classifiers

Vercel restored access to Anthropic’s Claude Fable 5 on AI Gateway after US export controls were lifted, with updated safety classifiers now in place. Routine tasks may still trigger classifiers, so m…

Vercel launches Service Bindings for secure internal service communication

Vercel introduced Service Bindings to enable secure, internal communication between services in the same deployment. Services can now call each other via injected environment variables without exposin…

Resend joins Vercel Marketplace for email delivery

Resend is now available on the Vercel Marketplace, enabling Vercel teams to send transactional and marketing emails via API or SMTP without managing infrastructure. Developers can build emails as Reac…

Vercel launches Security Dashboard in private beta to flag misconfigurations

Vercel introduced a Security Dashboard in private beta to aggregate and flag security misconfigurations across accounts and projects, such as missing 2FA, public preview environments, or long-lived cr…

Vercel open-sources konsistent CLI linter for TypeScript codebases

Vercel released konsistent, an open-source CLI linter for TypeScript that enforces structural code conventions to ensure consistency for both AI agents and human developers. The tool is already used i…

Vercel CLI adds dry-run deployments for pre-deployment checks

Vercel CLI now supports dry-run deployments via `vercel deploy --dry`, allowing users to preview framework presets, included/ignored files, and file manifests before actual deployment. The feature out…