API Breaking Changes Radar

Shopify storefronts now support UCP 2026-08-25

Shopify storefronts now support UCP 2026-08-25

GPT 6 Astra from OpenAI is now available on Vercel AI Gateway for coding, computer use, research, and multi-step professional workflows.

Ling 3.0 Flash Sante is now available on AI Gateway for free
Ling 3.0 Flash Sante from Inclusion AI is free on Vercel AI Gateway through October 4 through a -free model ID or regular ID.
GPT 6 Astra now available on Vercel AI Gateway
GPT 6 Astra from OpenAI is now available on Vercel AI Gateway for coding, computer use, research, and multi-step professional workflows.
Zachary Chen, Kevin Dawkins, and 1 other

In Spain and France, an incoming call intended for a non-Spanish or non-French destination that is then forwarded internationally while presenting a Spanish or French domestic caller ID is now blocked

Twilio Elastic SIP Trunking logo
International Call Forwarding - Compliance Blocking for Spain & France 
 In Spain and France, an incoming call intended for a non-Spanish or non-French destination that is then forwarded internationally while presenting a Spanish or French domestic caller ID is now blocked, in line with local regulatory requirements. 

When you deploy a Worker, Wrangler bundles your code and compresses it before uploading. Previously, Cloudflare checked that compressed size and rejected deploys over 3 MB (Free) or 10 MB (Paid). That

Deploy larger Workers — up to 64 MiB for both free and paid plans
You can now deploy Workers with larger dependencies, heavier frameworks, and more code without hitting size limits.
When you deploy a Worker, Wrangler bundles your code and compresses it before uploading. Previously, Cloudflare checked that compressed size and rejected deploys over 3 MB (Free) or 10 MB (Paid). That limit has been removed. Cloudflare now only checks the uncompressed size of your bundle, which is 64 MiB across all plans.
To check your Worker's bundle size before deploying:
 value is your uncompressed bundle size. This is what counts against the 64 MiB limit. The 
 value is shown for reference but is no longer a limit.
Worker size limits documentation
Python Workers now support WSGI web frameworks like Django and Flask
Python web frameworks following the 
Web Server Gateway Interface (WSGI)
Asynchronous Server Gateway Interface (ASGI)
 specification can now be used in Python Workers.
Using web frameworks with Python Workers
Based on the web framework you are using, you can use either 
For WSGI frameworks like Django or Flask:
 is equivalent to creating a 
 method. If you want more control over the 
For 

Manage resources as code with ant apply

, which creates and updates agents, environments, skills, memory stores, and deployments from files in your repository. Describe each resource in a file, run 
, and approve the plan it prints. Commit the 
 lockfile it writes so that later runs, on your machine or in CI, update the same resources instead of creating new ones. See 
Manage resources as code with ant apply

Basic build machines are now available on Pro and Enterprise

Cursor Cloud Agents can now run in Vercel Sandbox
Run Cursor Cloud Agents on infrastructure you control with Vercel Sandbox, using scale-to-zero workers, isolated microVMs, and durable orchestration.
Member of Technical Staff
Basic build machines are now available on Pro and Enterprise
Pro and Enterprise teams can now select Basic build machines, 2 vCPU and 8 GB, a lower-cost option for smaller apps. New projects still default to Elastic.

Added new controls for long-running work with GPT-6 Astra in the Responses API:

, our most capable model, built for the hardest end-to-end work.
Use GPT-6 Astra for reasoning, coding, computer use, research, and document creation. It combines these capabilities to carry complex tasks from an initial request to a finished result, using the context and tools you provide.
Key changes to consider when migrating:
GPT-6 Astra does not support the 
GPT-6 Astra does not support custom 
 values or log probabilities (
Tool calling requires the Responses API. If you use tools with Chat Completions, follow the 
Responses migration guide
 asynchronously checks for potential issues during agent work in supported Responses API requests. Checks can trigger safety alerts or stop a conversation for review.
 for capabilities, prompting, and migration guidance. Explore 
 for browser and desktop workflows, and see 
 for available inference tiers.
Added new controls for long-running work with GPT-6 Astra in the Responses API:
: Let the model continue working while your application runs function or custom tools, then return results as they become available.
: Send additional instructions while a response is in progress over WebSockets, so the model can incorporate corrections or cha

If some routes in a batch fail (for example, an invalid CIDR), the routes that were created successfully are removed from the form automatically, so you only need to fix and resubmit the ones that fai

Define custom applications for breakout and prioritized traffic from the Cloudflare One Appliance dashboard
 directly from the dashboard, without calling the API.
Assign application traffic
 to create a custom application matched by 
 field, alongside Cloudflare-managed applications.
Edit or delete an existing custom application from the same panel, no API round-trip required.
 lets you match traffic by its source IP range, complementing the existing 
source LAN interface breakout criteria
This complements the existing API and Terraform workflow for managing applications.
Configure DHCP options from the dashboard on Cloudflare One Appliance
 directly from the dashboard when the 
 is acting as the DHCP server for a LAN.
 to choose from common options for PXE / iPXE boot, VoIP phone provisioning, and vendor-specific configuration, or select 
 to enter your own option code, type, and value.
This complements the existing 
API and Terraform workflow
 for configuring DHCP options.
New in Images: text rasterization and updates to the binding
We've added more ways to manage and manipulate images with the 
Render text into an image.
 Output a string of text into its own image or draw it ove

Adds MB WAY support to the hosted invoice page

Adds MB WAY support to the hosted invoice page

Text generated by Claude Fable 5.1 and Claude Mythos 5.1 carries Anthropic's text watermark, and supported image, video, and audio files that Claude produces through the

Text generated by Claude Fable 5.1 and Claude Mythos 5.1 carries Anthropic's text watermark, and supported image, video, and audio files that Claude produces through the 
The guides for the Claude Enterprise endpoints of the 
Claude Enterprise Analytics API
 header; send it on every request to these endpoints, as in the rest of the Claude API. See 

Muse Spark 1.3 now available on AI Gateway

Free domain with Pro offer now includes .app and .dev
The free domain with Pro offer now covers .app and .dev, bringing the eligible TLDs to eight. Every Pro team can claim one domain, free for the first year.
Jeremy Dopkin, Elliot Dauber, and 1 other
Muse Spark 1.3 now available on AI Gateway
Call Meta's Muse Spark 1.3 through Vercel AI Gateway with one API key: no provider account, fallbacks, spend tracking, and traces on every request.
GLM-5.3 is 50% off through DigitalOcean on AI Gateway
GLM-5.3 served by DigitalOcean is 50% off on Vercel AI Gateway through September 7. Call the promo model name or use provider options to get the discounted rate.
Gemini 3.8 Flash now available on AI Gateway
Call Google's Gemini 3.8 Flash through Vercel AI Gateway with one API key: no provider account, fallbacks, spend tracking, and traces on every request.
Rohan Taneja, Zachary Chen, and 1 other

Upgrade to Java Helper Library 13.0.0 or later as soon as possible. Version 13 does include breaking changes so follow the

Verify Custom Templates are now Generally Available
 give you the flexibility and security to use custom SMS/RCS and Voice message templates beyond standard 
defaults or pre-approved options
To learn more about how to create a custom template for your business needs, check out our 
Critical fix: Java Helper Library 13.0.0: fixes crash on unknown fields in nested API responses
The Java Helper Library now handles unknown fields gracefully in nested response objects, but if customers are on an affected version, their integration is at risk until they upgrade.
Versions 10.0.0 through 12.1.1 parse nested response types (profile, configuration, webhook, and others) with strict deserialization. When an API response comes back with a field the SDK doesn't recognize, the request fails with ApiConnectionException: Unable to deserialize response: Unrecognized field. Because Twilio APIs can add new optional fields at any time, this is a latent break. Customers’ integrations work today and can start failing without warning the moment a new field ships.
In 13.0.0, the Java Helper Library ignores unknown fields in all response objects, matching how every other Twilio Helper Library already behave

Updated API errors so applications can distinguish traffic that increases too quickly from temporary model overload.

Updated API errors so applications can distinguish traffic that increases too quickly from temporary model overload.
Traffic that increases too quickly can return a 
 code. Temporary model overload returns a 
 code. Both responses may include 
. When the header is present, wait at least as long as it specifies before retrying. If it's missing, use exponential backoff. See the 

AWS PrivateLink is now available on Pro and Enterprise

Qwen 3.8 Max 0902 now available on AI Gateway
Call Alibaba's Qwen 3.8 Max 0902 through Vercel AI Gateway with one API key: no provider account, fallbacks, spend tracking, and traces on every request.
Claude Fable 5.1 now available on AI Gateway
Call Anthropic's Claude Fable 5.1 through Vercel AI Gateway with one API key: no provider account, fallbacks, spend tracking, and traces on every request.
Rohan Taneja, Jerilyn Zheng
AWS PrivateLink is now available on Pro and Enterprise
AWS PrivateLink is now generally available on Vercel. Connect deployments to AWS-hosted databases and services over a private network path.
Miroslav Simulcik, Lakshay Bhushan, and 4 others

As of July 2026, we require that apps submitting to the Slack Marketplace have at least 10 installations on active workspaces, up from the temporary lower threshold. Your app needs to stay at 10 or mo

Updated installation requirement for Slack Marketplace apps
Updated installation requirement for Slack Marketplace apps
As of July 2026, we require that apps submitting to the Slack Marketplace have at least 10 installations on active workspaces, up from the temporary lower threshold. Your app needs to stay at 10 or more installed workspaces throughout the 
entire duration of the review

AI Gateway monthly usage invoices, issued at the beginning of each month for the previous month's usage, now show a single total cost for each model. These invoices no longer break out input and outpu

AI Gateway consolidates monthly usage invoice line items and standardizes model names
AI Gateway monthly usage invoices, issued at the beginning of each month for the previous month's usage, now show a single total cost for each model. These invoices no longer break out input and output token quantities and unit prices into separate line items. This change does not apply to invoices for AI Gateway credit purchases.
For example, an invoice that previously included these separate line items:
anthropic claude-haiku-4-5-20251001 Input Tokens
: 40,000 tokens at $0.000001 ($0.04)
anthropic claude-haiku-4-5-20251001 Output Tokens
: 24,000 tokens at $0.000005 ($0.12)
The updated invoice includes one line item: 
anthropic/claude-haiku-4.5
AI Gateway has also standardized model names across invoices and logs. Model variants that previously appeared with provider-specific version suffixes now use a consistent 
Unified Billing documentation
AI Gateway logging documentation

Thinking blocks produced by Claude Fable 5.1 and Claude Mythos 5.1 are preserved only for the model that produced them or a newer one: earlier models can't read them, and the API drops one replayed to

), the successor to Claude Fable 5 for long-running agentic coding, knowledge work, and research, alongside 
, at $10 / $50 USD per MTok, the same as Claude Fable 5, with cache reads cut to $0.25 per MTok. Claude Fable 5.1 is available on the Claude API, 
What's new in Claude Fable 5.1
 for capabilities, API changes, and migration guidance.
Prompt cache reads on Claude Fable 5.1 and Claude Mythos 5.1 cost $0.25 USD per million tokens: 0.025x the base input price, compared with 0.1x on other models. Cache writes are unchanged. See 
On Claude Fable 5.1 and Claude Mythos 5.1, 
 aren't supported and return a 400 error. 
 are unchanged. To guarantee schema-conformant tool inputs, use 
Thinking blocks produced by Claude Fable 5.1 and Claude Mythos 5.1 are preserved only for the model that produced them or a newer one: earlier models can't read them, and the API drops one replayed to an earlier model. Claude Fable 5.1 accepts thinking blocks from Claude Opus 5, Claude Fable 5, Claude Mythos 5, and earlier Claude models. On Claude Fable 5.1, the API also 
checks that nothing before a block has changed
: for new accounts created on or after August 31, 2026, replaying one after the 
, or an 

The Polaris CDN is adopting semantic versioning

Polaris CDN 1.1 release candidate
The Polaris CDN is adopting semantic versioning

AI Gateway budgets now support a user scope. Cap what an individual team member spends through the gateway, alongside team, project, and API key budgets.

fx is now available in the AI SDK harness layer
fx is now available in the AI SDK harness layer through the @ai-sdk/harness-fx adapter, built on the ACP harness. Run it through the same HarnessAgent interface.
Set per-user budgets on AI Gateway
AI Gateway budgets now support a user scope. Cap what an individual team member spends through the gateway, alongside team, project, and API key budgets.
Jeremy Philemon, Sam Chitgopekar, and 3 others
Vercel Sandbox now calculates snapshot storage costs daily
Snapshot Storage is now calculated daily, so every charge maps to the storage you held that day. Pricing is unchanged at $0.08 per GB-month.
Christian Le, Pranav Kanchi, and 2 others

with access controlled by your organization's roles and permissions.

 User lifecycle management 
X.509 workload identity federation
 are now generally available for the OpenAI API. Configure certificates and X.509 identity providers directly in the 
, with access controlled by your organization's roles and permissions.

We've fixed the help template so hidden subcommands are no longer listed under their parent command.

Release: Slack CLI v4.7.0
Release: Slack CLI v4.7.0
 of the Slack CLI has been released!
 command, which checks the status of your most recent request to have an app approved for install.
We've added support for custom app icons. The 
 commands can now be used to upload an icon file.
We've fixed the help template so hidden subcommands are no longer listed under their parent command.

Your account has exceeded D1's free tier daily row read limit. Upgrade to a paid plan or wait until tomorrow (midnight UTC) to continue.

D1 enforces free tier daily query limits
Beginning September 1, 2026, D1 queries on the 
 will fail when an account exceeds the daily 
row read or row write limits
 will return errors until the limit resets at midnight UTC. Stored data is not affected.
You will receive email alerts when the daily limit is reached. The following errors indicate that a limit has been exceeded:
Your account has exceeded D1's free tier daily row read limit. Upgrade to a paid plan or wait until tomorrow (midnight UTC) to continue.
The account has reached its daily row read limit.
Your account has exceeded D1's free tier daily row write limit. Upgrade to a paid plan or wait until tomorrow (midnight UTC) to continue.
The account has reached its daily row write limit.
Inspect database query activity before the enforcement date to identify queries that may exceed these limits. To reduce row reads, add 
 to tables and review queries that perform full table scans. If usage requires higher limits after optimization, upgrade to a 
For more information on D1 errors and how to handle them, refer to the 
Crawl endpoint now respects the Content Signals `use` directive
 endpoint now respects the 
 standard, letting 

Charges now include a sublist of refunds through the

Charges now include a sublist of refunds through the 

no beta header is required. It's available on Claude Fable 5, Claude Mythos 5, Claude Opus 4.5 and newer, and Claude Sonnet 4.5 and newer; see the code execution tool's

; no beta header is required. It's available on Claude Fable 5, Claude Mythos 5, Claude Opus 4.5 and newer, and Claude Sonnet 4.5 and newer; see the code execution tool's 

MiniMax H3 video generation is 50% off on Vercel AI Gateway through September 13. This applies to both H3 and H3 Max. The model IDs are unchanged, so requests you already send get the discount.

MiniMax H3 and H3 Max are 50% off on AI Gateway
MiniMax H3 video generation is 50% off on Vercel AI Gateway through September 13. This applies to both H3 and H3 Max. The model IDs are unchanged, so requests you already send get the discount.
Josh Lipman, Jerilyn Zheng

The Assistants API shut down on August 26, 2026. Migrate to the Responses API and Conversations API using the

gpt-4o-transcribe-diarize
Announced the deprecation of 
gpt-4o-transcribe-diarize
. These models will shut down on February 26, 2027. Migrate to 
The Assistants API shut down on August 26, 2026. Migrate to the Responses API and Conversations API using the 

More resilient refreshes for expiring offline access tokens

More resilient refreshes for expiring offline access tokens

Hy4 Preview now available on AI Gateway

Hy4 Preview now available on AI Gateway
Call Tencent's Hy4 Preview through Vercel AI Gateway with one API key: no provider account, automatic fallbacks, spend tracking, and traces on every request.
Build and deploy eve agents from the Vercel dashboard
Build eve agents from the Vercel dashboard. The builder scaffolds the code, pushes it to your Git repo, and creates a Vercel project with AI Gateway models, web chat or Slack, and MCP connections.
Vercel CLI expands commands for DNS, domains, and projects
New Vercel CLI commands bring more DNS, domain, and project workflows to the terminal. Inspect and update DNS records, renew domains, configure projects, and manage members interactively or through automation.
David Lee, Melkey Moksyakov

Meet Rosalind Workbench: Empowering every scientist to be their own research team

 ChatGPT Work local security 
 ChatGPT Work: usage and cost 
 Meet Rosalind Workbench: Empowering every scientist to be their own research team 

A new GA release for the macOS Cloudflare One Client is now available on the

AI Search now supports GLM-5.3 Flash
 for text generation. The model has a 1,048,576-token context window and runs on Workers AI.
To configure the model for an AI Search instance, refer to 
Cloudflare One Client for macOS (version 2026.7.1376.0)
A new GA release for the macOS Cloudflare One Client is now available on the 
stable releases downloads page
This hotfix resolves an issue where a small but noticeable percentage of DNS queries fail across platforms.
Cloudflare One Client for Windows (version 2026.7.1376.0)
A new GA release for the Windows Cloudflare One Client is now available on the 
stable releases downloads page
Fixed a rare but critical issue where the client could fail to connect or switch organizations due to an invalid registration after switching installed client versions. Additionally, this hotfix resolves an issue where a small but noticeable percentage of DNS queries fail across platforms.
Cloudflare One Client for Linux (version 2026.7.1377.0)
A new GA release for the Linux Cloudflare One Client is now available on the 
stable releases downloads page
This hotfix resolves an issue where a small but noticeable percentage of DNS queries fail across platforms.
Z.ai

Password Protected Shop Dev flows on Shopify Theme CLI v3.83.x and older to be deprecated

Password Protected Shop Dev flows on Shopify Theme CLI v3.83.x and older to be deprecated
Build Shopify apps in PHP and Python with new official packages
Oxygen is now available on trial plan stores

Cursor is now available in the AI SDK harness layer

Run Claude Managed Agents with Chat SDK
Ship a Slack research bot built on Claude Managed Agents and Chat SDK. One persistent session per thread, streamed briefs with sources, and no Slack credentials to manage.
Cursor is now available in the AI SDK harness layer
Cursor now runs through the AI SDK harness layer via @ai-sdk/harness-cursor, using the same HarnessAgent interface as every other supported harness.
Find deployments faster with redesigned filters
The filters on the Deployments page have been redesigned with one-click suggestions, typeahead matching, and natural language queries to help you find deployments faster.
wits, Christopher Skillicorn, and 1 other
Ling 3.0 Flash Fin now available on AI Gateway for free
Ling 3.0 Flash Fin from Inclusion AI is free on AI Gateway through September 25, with a -free model ID that stops serving instead of billing.

Twilio SendGrid has added automatic SSL certificate provisioning as a toggle within the standard link branding setup flow. When enabled, SendGrid provisions and hosts the SSL certificate needed to ser

Update to Twilio SendGrid Link Branding: Automatic SSL Certificate Provisioning
Twilio SendGrid has added automatic SSL certificate provisioning as a toggle within the standard link branding setup flow. When enabled, SendGrid provisions and hosts the SSL certificate needed to serve tracking and branded links over HTTPS, removing the need to purchase, install, and maintain a certificate through an external CDN or DNS provider.
Twilio SendGrid recommends activating this setting for all link branding domains. Gmail is rolling out warnings on non-HTTPS links by the end of October 2026, and tracking links without SSL will see reduced click-through as a result. Enabling automatic SSL requires adding two additional DNS records provided by SendGrid, alongside the CNAME records already used for link branding.
The Link Branding API now supports a new auto_ssl parameter when creating a branded link. Setting auto_ssl to true provisions the certificate automatically, returns the additional DNS records needed to complete setup, and reports certificate status through a new auto_ssl_available field once the certificate is issued and live. auto_ssl can only be set when a link is created and can't b

Existing service token secrets continue to work and do not require rotation. Both formats use the same Client ID and the same

Access service token secrets use a scannable format
Cloudflare Access service token Client Secrets created on or after August 26, 2026, use the format 
cfast_[40 alphanumeric characters][8-character checksum]
. The prefix and checksum make these credentials easier for secret scanning tools to identify with fewer false positives.
Existing service token secrets continue to work and do not require rotation. Both formats use the same Client ID and the same 

with the same permissions, and stop working when the linked account is removed from an organization. This lets organization admins more easily track usage for each account, and ensure key usage is leg

In Python SDK 1.2.0, TypeScript SDK 0.122.0, Go SDK 1.68.0, Java SDK 2.59.0, Ruby SDK 1.67.0, and C# SDK 12.44.0, 
 beta headers and return the same shapes as 
client.beta.skills.delete()
 deletes a Skill together with all of its versions, and the beta Messages type 
 (the container Skill reference) is renamed 
. Requests that still send the beta headers keep receiving the beta shapes. See 
 in the Claude Console. They act as you or as a 
, with the same permissions, and stop working when the linked account is removed from an organization. This lets organization admins more easily track usage for each account, and ensure key usage is legitimate. These API keys can be scoped to a specific workspace or 
work on admin endpoints and across any workspace
 the account has access to. Workspace API keys remain supported as a legacy option. See 
To move an existing integration off the header, see 
To move an existing integration off the header, see 

Four additional topics are now available for Events

Four additional topics are now available for Events
GitHub commits now name the last theme editor

Muse Image now available on AI Gateway

Muse Image now available on AI Gateway
Call Meta's Muse Image through Vercel AI Gateway with one API key: no provider account, automatic fallbacks, spend tracking, and traces on every request.
Josh Lipman, Zachary Chen, and 1 other
Gemini 3.5 Transcribe now available on AI Gateway
Call Google's Gemini 3.5 Transcribe through Vercel AI Gateway with one API key: no provider account, automatic fallbacks, spend tracking, and request traces.
Kevin Dawkins, Jerilyn Zheng
Qwen 3.8 Flash now available on AI Gateway
Call Alibaba's Qwen3.8-Flash through Vercel AI Gateway with one API key: no provider account, automatic fallbacks, spend tracking, and request traces.
GLM 5.3 Flash now available on AI Gateway
Call Z.ai's GLM-5.3 Flash through Vercel AI Gateway with one API key: no provider account, automatic fallbacks, spend tracking, and traces on every request.
Vercel Security Dashboard is now generally available
The Vercel Security Dashboard gives you a single place to view the security posture of every project.
Roman Olynyk, Kostyantyn Voytenko, and 2 others
Python projects now support routing rules
Set response headers and rewrite requests for FastAPI, Django, and Flask apps with routing ru

Twilio has observed  a sudden rise in SMS Pumping fraud (aka SMS Artificially Inflated Traffic) targeting messages to Italy (IT) and Argentina (AR). To protect customers from this activity, Twilio Ver

Verify tightens fraud prevention geo permissions for Italy and Argentina
Twilio has observed  a sudden rise in SMS Pumping fraud (aka SMS Artificially Inflated Traffic) targeting messages to Italy (IT) and Argentina (AR). To protect customers from this activity, Twilio Verify is tightening its fraud prevention posture for these two countries, effective August 19, 2026.
Specifically, Twilio is  updating the default Verify Geo Permissions setting for Italy and Argentina from "Allow all traffic" — the default state at account creation — to "Monitor all traffic for fraud." This change applies to all existing  accounts where IT and/or AR remain in  the default "Allow all traffic" state.
What this means for customers
If your account still has IT and/or AR set to the default 
 permission, it will be updated to 
"Monitor all traffic for fraud"
 starting August 19, 2026.
If you've already changed the geo permission for IT or AR away from the default “Allow all traffic” (to any other state), your configuration will not be modified by this update.
The protection provided by "Monitor all traffic for fraud" requires that you have Fraud Guard enabled. 
No action is required for customers with Fr

Microsoft is deprecating the Azure Monitor HTTP Data Collector API. Support for the API ends on 2026-09-14. As a result, Cloudflare will no longer maintain the Azure Functions-based connector after th

New Workers AI text generation models in AI Search
 now supports six additional 
 models for text generation:
@cf/deepseek-ai/deepseek-v4-flash-0731
@cf/deepseek-ai/deepseek-v4-pro-0813
@cf/moonshotai/kimi-k2.7-code
These models run on Workers AI, so they do not require an additional provider key. Select a model when creating or updating an AI Search instance in the dashboard or through the API.
For the full list of supported models, refer to 
Delete Log Explorer datasets
Cloudflare Log Explorer customers can now permanently delete account and zone datasets from the Cloudflare dashboard or API.
Deletion protection is enabled by default to prevent accidental data loss. In the dashboard, go to 
, disable deletion protection for the dataset, select 
, and enter the dataset name to confirm.
To delete a dataset through the API, first set 
Update an account or zone dataset
Delete an account or zone dataset
Dataset deletion is irreversible and runs asynchronously. You cannot recreate the same dataset while deletion is in progress.
Azure Functions-based Microsoft Sentinel connector deprecation
Cloudflare Enterprise customers using the 
Azure Functions-based Microsoft Sentinel connector
Clo

Adds the ability to query entitlements and launch the customer portal in Customer Sessions

Adds support for updating Connect parameters of an existing Payment Link
Adds a funding source identifier to Link payments
Adds new error codes for payment method failures
Adds the ability to restrict card funding types in Checkout Sessions
Adds support for metadata in confirmation tokens
Adds a funding source identifier to Link wallet card payments
Adds a customer update deep link for billing portal sessions
Adds the ability to query entitlements and launch the customer portal in Customer Sessions
Adds frozen fields property to Invoice Items
Adds support for the Billie payment method for invoices and subscriptions
Increases the subscription item limit
Adds support for specifying place of supply scheme on IGIC tax in the EU

They cover organization info, members, invites, workspaces and workspace members, API keys, rate limits, service accounts, workload identity federation issuers and rules, and customer-managed encrypti

 session endpoints are out of beta for Cowork and Claude Code sessions. See 
Retrieve session transcripts
 local session endpoints now also return transcripts of Claude Science sessions (
) and Claude for Microsoft 365 sessions in Excel, PowerPoint, Word, and Outlook (
), in beta for Claude Enterprise organizations, with your existing Compliance Access Key and the 
 CLI and the Python, TypeScript, C#, Go, Java, PHP, and Ruby SDKs under 
. They cover organization info, members, invites, workspaces and workspace members, API keys, rate limits, service accounts, workload identity federation issuers and rules, and customer-managed encryption keys. Usage and cost reports and the Claude Enterprise user-management and analytics endpoints remain curl-only. The CLI and SDKs read an Admin API key from 
 for Claude Fable 5, Claude Mythos 5, Claude Opus 5, Claude Sonnet 5, and Claude Opus 4.8. Requests use the same 
 entries as on the Claude API.

Chat SDK's Slack adapter now supports Slack Enterprise Grid.

Vercel applications are protected from Next.js August 2026 security vulnerabilities
Applications hosted on Vercel are protected from recently disclosed Next.js vulnerabilities.
Steven Salat, Zack Tanner, and 1 other
Speed Insights now has a free tier
A free performance overview from real users, now on every plan, with paid capabilities continuing as Speed Insights Plus.
Damien Simonin Feugas, Chris Widmaier
MiniMax M3 and M2.7 are free on AI Gateway
MiniMax M3 and M2.7 are free on Vercel AI Gateway through September 6 when GMI Cloud serves them, using the -free model IDs or a provider pin.
Zachary Chen, Jerilyn Zheng
Wan 3.0 now available on AI Gateway
Alibaba's Wan 3.0 is now on Vercel AI Gateway, generating up to 30 seconds of video with audio from text, image, video, and audio references.
Josh Lipman, Jerilyn Zheng
AI Gateway now supports asynchronous video generation
AI Gateway can now run video generation as a background job, with three ways to wait for it: a webhook, a start and status pair, or short polling requests.
Kevin Dawkins, Gregor Martynus, and 3 others
Vercel Connect is now generally available
Vercel Connect, the secure connectivity layer for AI agents and apps, is 

Automating repetitive work at OpenAI with Codex

 Automating repetitive work at OpenAI with Codex 
 Meet the winners of OpenAI Build Week 

Cloudflare Access administrators can now choose a grace period when rotating a service token secret. Both secrets remain valid during the grace period, giving administrators time to update services wi

Create app-scoped API tokens for Flagship
. These tokens grant access only to the Flagship apps you select, instead of every app in the account.
When you create a custom token, open the resource dropdown (it defaults to 
. Then choose the app and a 
 permission: Evaluate, Read, or Write. Account-wide Flagship Evaluate, Read, and Write permissions still exist when you need access to every app.
Use app-scoped tokens in trusted server-side environments, such as Wrangler, CI, or a backend service that should only touch one app.
To create a token, refer to 
open the app-scoped token form
Grace periods for service token rotation
Cloudflare Access administrators can now choose a grace period when rotating a service token secret. Both secrets remain valid during the grace period, giving administrators time to update services without interrupting authentication.
The dashboard offers grace periods from one hour to 30 days. Administrators can also revoke the previous secret immediately. The API accepts an RFC 3339 expiration time for custom rotation schedules.
For configuration instructions, refer to 
Rotate service token secrets
Temporarily turn off Access service tokens
Cloudflare Access ad

Script tags are deprecated and will stop running on March 1, 2027

Script tags are deprecated and will stop running on March 1, 2027

Vercel Functions using the Bun runtime now support larger package sizes up to 5GB and extended max duration up to 30 minutes.

Elastic build machines now use Turborepo cache hits to prevent downgrades
Elastic build machines now use Turborepo cache data when deciding tier downgrades, keeping heavily cached builds on their machine tier to prevent underprovisioning.
Vercel Sandbox is now globally available
Vercel Sandbox now runs in iad1, sfo1, cle1, and cdg1. Choose a region per sandbox, set a project default, and configure failover regions.
Marc Codina Segura, Tom Lienard, and 3 others
Environment variables now use Config and Secret types
Choose Config or Secret when adding environment variables in the dashboard and CLI. Config values stay readable to authorized members; Secret values are write-only after saving.
Bun runtime now supports large functions and extended max duration
Vercel Functions using the Bun runtime now support larger package sizes up to 5GB and extended max duration up to 30 minutes.

Dashboard SCIM now supports replacing groups using HTTP

Download the Cloudflare One Virtual Appliance for your hypervisor from the dashboard
Cloudflare One Virtual Appliance
, you can now select your hypervisor and download the appliance directly from the dashboard — no need to look up asset URLs.
, then select your hypervisor: 
Download the OVA image (VMware ESXi) or the install script (Proxmox and libvirt/KVM) for the selected hypervisor.
 to open deployment instructions for your platform.
This complements the existing self-serve 
registration and license key generation
Configure a Cloudflare One Virtual Appliance
Test Data Loss Prevention profiles without sending traffic through Gateway
Data Loss Prevention (DLP)
 evaluates sample content before you apply a profile to production traffic. Paste text, upload a file, or upload a HAR file, then select the profiles you want to test.
Test scan sends content directly to the DLP scanner. Gateway policies are not evaluated, no traffic passes through Gateway, and no Gateway activity logs are created. Results include matched profiles, detection entries, confidence levels, match context, proximity keywords, file metadata, antivirus status, and OCR output.
Test scan is available to all Cloudflare

The 1-hour cache duration for prompt caching no longer requires a beta header. Learn more in

. The SDK's HTTP layer moves from 
, a maintained, API-compatible fork: build custom 
, and transport objects from 
 helpers are unchanged), and call 
 at startup if you rely on tracing or mocking libraries that patch 
. v1.0 requires Python 3.10 or later and removes long-deprecated surface, including the legacy Text Completions API, the 
 parameters on Messages methods, and the tool runner's client-side 
 now raises an error when no AWS region is configured instead of defaulting to 
 for every change with before-and-after snippets.
 is out of beta on the Claude API as the 
 is out of beta on the Claude API. Requests to the 
 beta header. Requests sent without the header use the current response format: 
) are out of beta on the Claude API. Requests no longer require the 
 (claude.ai) organizations (members, invites, groups, and custom roles) are out of beta. The 
), our most capable widely released model. Claude Opus 4.8 supports a 
. Authenticate workloads to the Claude API with short-lived OIDC tokens from your own identity provider (AWS IAM, Google Cloud, GitHub Actions, Kubernetes, Microsoft Entra ID, Okta, SPIFFE, and more) instead of long-lived static API keys. Configure iss

Shopify Analytics is now a full-stack platform for apps. Model, query, embed, and enrich analytics on Shopify without managing your own data infrastructure.

Full-stack capabilities to power app analytics
Shopify Analytics is now a full-stack platform for apps. Model, query, embed, and enrich analytics on Shopify without managing your own data infrastructure.

DeepSeek V4 Flash Vision Experimental now available on AI Gateway

Deployment Storage keeps your deployments rollback-ready
Deployment Storage is now measured and billed at $0.10 per GB per month for new Pro and Enterprise teams. Control retention in Deployment History.
Jay Gengelbach, Mehul Kar, and 2 others
Vercel CLI expands support for DNS, domains, and project commands
New Vercel CLI commands bring more DNS, domain, and project workflows to the terminal. Inspect and update DNS records, renew domains, configure projects, and manage members interactively or through automation.
David Lee, Melkey Moksyakov
GPT-5.6 Sol is now 50% off a lower price
GPT-5.6 Sol list pricing drops to $4 and $20 per million tokens, and the 50% AI Gateway discount still applies, so requests now cost even less.
DeepSeek V4 Flash Vision Experimental now available on AI Gateway
Call DeepSeek V4 Flash with image input through Vercel AI Gateway with one API key, a 1M token context window, fallbacks, and a trace on every request.
Kevin Dawkins, Jerilyn Zheng
Connect v0 apps to Slack, Google, and 100+ other services
Apps and agents built in v0 can now securely connect to 100+ third-party services, including Slack, Google, Notion, GitHub, and Salesforce, through Vercel Connect

API customers can now select regional processing for an individual request by using a prefixed domain with an API key from a project having Global geography. Existing eligibility, data retention contr

 Scaling cyber defenders with Daybreak 
API customers can now select regional processing for an individual request by using a prefixed domain with an API key from a project having Global geography. Existing eligibility, data retention control, endpoint, and model support requirements continue to apply. Learn more in the 
GPT-5.6 Sol now costs $4 per million input tokens and $20 per million output tokens, representing 20% lower input pricing and 33% lower output pricing. GPT-5.6 Sol’s promotional pricing is available at least through November 21, 2026. See 

Up to five login profiles can be saved per device, and saved profiles can be removed from the profile list at any time.

Choose OAuth scopes for Wrangler and the Cloudflare API MCP server
Cloudflare API MCP server
 now use optional OAuth scopes. During authorization, you can choose which optional scopes to grant instead of approving every scope requested by each client.
The consent dialog now includes the option to edit the permissions you grant to Wrangler or the Cloudflare API MCP server:
You can then choose which specific permissions to grant:
Required scopes remain selected. Choosing fewer optional scopes limits each tool's access to the permissions needed for your workflow.
If a command or tool call needs a scope that you declined, reauthorize the client and grant that scope.
Edit optional permissions
Automatically remediate Microsoft 365 and Google Workspace findings with API-based CASB remediation policies
 is an API-based (agentless) tool that continuously scans your SaaS and cloud applications for security misconfigurations and data exposure. You can now use 
CASB remediation policies
 to automatically fix a finding or send a webhook the moment CASB detects it, without manual triage.
Remediate Microsoft 365 and Google Workspace findings
A policy can perform a first-party remediation action d

Backup scheduling now retries on failure, anchored to its original scheduled window so a retry can't skip or re-scan the wrong projects, reads eligibility from the primary database instead of a replic

Read replicas moved to Project Settings → Infrastructure
Read replica management now lives on 
Project Settings → Infrastructure
, next to compute and disk. 
 is for Pipelines destinations only.
Creating, dropping, and managing replicas
Read replica docs (updated for the new location)
What changes in the Dashboard:
Project Settings → Infrastructure
 to list replicas, open the topology, and add a replica
Old replica detail URLs redirect to Infrastructure
?destinationType=Read+Replica
 on Replication open the Infrastructure add-replica flow
Replication shows a short note pointing to Infrastructure for anyone who still opens that page looking for replicas
 for the updated create path.
Fixed daily backups occasionally skipped due to a scheduling timeout
The job that schedules daily backups checks which projects are due on a strict time budget. Under database replica lag, that check could exceed its budget, and with no retry in place the whole 10-minute scheduling window was silently dropped fleet-wide — any project due in that window missed its backup for the day.
Backup scheduling now retries on failure, anchored to its original scheduled window so a retry can't skip or re-scan the wr

), a client toolset for driving a browser that your application hosts. It works inside a browser viewport rather than a whole desktop, reading the page itself (its accessibility tree, elements, forms,

The Claude Platform release notes list changes to the Claude API, the client SDKs, and the Claude Console, newest first.
 is now generally available on the Claude API as the 
computer_toolset_20260801
 toolset: no beta header, batch actions (several actions in one turn), 
 enabled by default, and per-member configuration through 
. Earlier beta versions remain available. Upgrading an existing integration changes the request shape and tool handling; see 
), a client toolset for driving a browser that your application hosts. It works inside a browser viewport rather than a whole desktop, reading the page itself (its accessibility tree, elements, forms, and tabs) and adding element references, form input, tab management, download reporting, and opt-in file upload on top of screenshot-and-click control.
Both toolsets are available for Claude Fable 5, Claude Mythos 5, Claude Opus 5, Claude Sonnet 5, and Claude Opus 4.8 on the Claude API.

Checkout And Accounts Configuration API for unified branding across checkout, customer accounts, and sign-in

Checkout And Accounts Configuration API for unified branding across checkout, customer accounts, and sign-in

Vercel Functions now support the new Bun 1.4 release, a complete rewrite of Bun from Zig to Rust, with many improvements to Node.js compatibility and bug fixes.

Manage Vercel Toolbar comments from the CLI
List, inspect, reply to, and resolve Vercel Toolbar comments from the terminal with the new vercel comments command.
Julian Benegas, Vishal Yathish, and 1 other
Custom metrics are now supported in Vercel Observability
Emit your own metrics from Vercel Functions and analyze them alongside Vercel's built-in observability data.
Vercel Agent is now available in Slack code channels
Start a code channel to give a task its own shared Vercel Agent session, where your team can send requests, review diffs, and coordinate together.
Hiroki Osame, Bryan Hunter, and 2 others
Manage Vercel Container Registry with Vercel CLI
Manage Vercel Container Registry with Vercel CLI. The new vercel vcr commands let you create and inspect repositories, browse images and tags, and connect your existing container tooling. Docker, Podman, and Buildah are all supported.
Marc Codina Segura, Andy Waller
Bun 1.4 is now available in Vercel Functions
Vercel Functions now support the new Bun 1.4 release, a complete rewrite of Bun from Zig to Rust, with many improvements to Node.js compatibility and bug fixes.
Florentin Eckl, Tom Lienard

Transparent backgrounds are now available in preview for

 on the OpenAI API platform. Track your cache hit rate over time, cache reads per write, and the breakdown of cache-read, cache-write, and uncached tokens to understand your caching efficiency and identify opportunities to improve. Filter metrics by model and service tier.
Transparent backgrounds are now available in preview for 
 in the Images API and the Responses API image generation tool. Set 
 does not support transparent backgrounds. Learn more in the 

Agent updates and the deprecation of

The next wave of Agent messaging experience
The next wave of Agent messaging experience
We're rolling out a wave of updates to agents on Slack, along with an important heads-up about the road ahead for the Assistant messaging experience.
Agent updates and the deprecation of 
The Assistant messaging experience (
. The assistant-related methods that power it are being replaced by a new set of API methods: the 
. Agent sessions let users manage their conversations with your agent in one place, with a status, a title, and a stop button. The 
agents.sessions.setStatus
assistant.threads.setTitle
 API methods, respectively. For now, your existing calls keep working through a compatibility bridge. See the 
The Agent messaging experience (
) is already available, and we recommend moving to it now. If your app still uses 
 to move to the new experience. Read more in the guide to 
Alongside this announcement, we're releasing a set of improvements to the agent experience:
: a newly renamed and revamped 
 tab helps users find popular agents with a new landing page.
Improved thread experience
: including multi-agent stacking, thread naming, a revamped thinking UI, and a native stop button for ag

Members with only resource-scoped Access roles can now open Access resource list pages in the Cloudflare dashboard and call list endpoints in the API. They no longer need an additional account-scoped

Enriched 403 responses for the Cloudflare API
 field that links directly to the API documentation for the endpoint that was denied. This gives developers, administrators, and agents an immediate path to the relevant docs with role information instead of guessing at which role or permission they are missing for that endpoint.
Enriched 403 error responses
: When a Cloudflare API request is denied, the error response now includes a 
 field that points to the documentation for that specific endpoint. Contextual 403 responses are now available across nearly all Cloudflare product APIs.
: The linked API docs surface the roles required for each endpoint, making it easier to self-serve access issues.
Better support for tools and agents
: Agents can use the \documentation_url` field to immediately fetch the endpoint's documentation from the 403 error response, identify the accepted permissions for the denied action, and use that context to drive third-party approval workflows.`
"https://developers.cloudflare.com/api/resources/workers/subresources/beta/subresources/workers/methods/list"
Browse the Cloudflare API documentation
Review API token permissions
Run more headless browsers concurrent

Hydrogen developer preview update: August 18, 2026

 now open as a full-page navigation
Hydrogen developer preview update: August 18, 2026