What changed in the Cloudflare API

This is a breaking change. @cloudflare/computer stores files in different tables from @cloudflare/shell, and there is no automatic data migration between workspace formats.

WAF Release - 2026-10-21 - Emergency - Cloudflare Managed Ruleset
This emergency release adds four rules to the Cloudflare Managed Ruleset to detect anomalous request patterns. Two rules block matching requests, and two log matches without blocking.
Generic Rules - Anomalous Execution - 1
Generic Rules - Anomalous Execution - 2
Generic Rules - Anomalous Execution - 3
Generic Rules - Anomalous Execution - 4
WAF Release - 2026-10-21 - Emergency - Cloudflare Free Ruleset
This emergency release adds two rules to the Cloudflare Free Ruleset to block anomalous request patterns.
Generic Rules - Anomalous Execution - 1
Generic Rules - Anomalous Execution - 2
Generic Rules - Anomalous Execution - 3
Generic Rules - Anomalous Execution - 4
Think agents now use a @cloudflare/computer workspace
 now use the workspace from 
 as well as the filesystem tools. A JavaScript execution environment can be configured by adding the 
 is deprecated, and so is Think's 
. These will no longer be developed nor supported from today onwards.
 { WorkerJavaScriptBackend } 
 "@cloudflare/computer/backends/worker-javascript"
 { WorkerJavaScriptBackend } 
 "@cloudflare/computer/backends/worker-javascript"
Migrate y

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

Cloudflare One Client for Windows (version 2026.8.2115.0)
This hotfix addresses an issue building trusted certificate chains.
Cloudflare One Client for macOS (version 2026.8.2115.0)
This hotfix addresses an issue building trusted certificate chains.
Cloudflare One Client for Linux (version 2026.8.2115.0)
A new GA release for the Linux Cloudflare One Client is now available on the 
This hotfix addresses an issue building trusted certificate chains.
Cloudflare API MCP server now supports step-up authorization
Grant only the required scopes when you connect, then authorize more on demand as your agent performs a task. When a tool call needs a scope you have not granted, the server returns an 
 naming it, and your MCP client asks you to approve it. Scopes you already granted are kept.
To start with the least access, choose 
 on the consent page when you connect 
Available now in all MCP clients that support step-up authorization.
Introducing Clairic: an extraction model for typed values, now on Workers AI
 is now available on Workers AI as a preview. Clairic is the first 
: a System One model that answers with values instead of text. We intend to release the model with open weights soo

Cloud Workload Security is now available in Beta

Precursor adds Monitor mode
We released Monitor mode at customers' request so they can test Precursor and the Precursor Interstitial without updating bot score.
 means Precursor detections can affect bot score and Challenge clearance. Selecting 
 means Precursor detections do not affect bot score or Challenge clearance.
We also separated the Precursor Interstitial from the mode. Set the sitewide default with 
 to customize it for specific requests. For example, leave 
 off across most of your site, but require an interstitial Challenge on 
 to enforce a valid session before purchase.
Cloud Workload Security is now available in Beta
Cloud Workload Security (CWS)
 is now available in Beta. Apply policy controls to cloud API requests before they reach your cloud provider.
Define policies with request filters and Rego checks. For example, block requests that would make an R2 bucket public. Use workloads and policy sets to target your cloud environments.
Manage your policies in the Cloudflare dashboard. Run the CWS proxy in your environment and route client requests through it to enforce your policies.
CWS requires a Zero Trust subscription. The Zero Trust Free plan is sufficient.
To co

The Local Explorer API exposes the same operation for tools and AI agents.

Test Cron Triggers with Local Explorer
Local Explorer can now send one-off Cron Trigger invocations to a Worker's 
 handler during local development. This removes the need to construct request URLs when testing handler behavior.
Start your local development server with 
 on the server URL and select 
The browser interface supports the following tasks:
Invoke configured cron expressions
Set reusable scheduled times
Ad hoc triggers and scheduled time presets stay in your browser. They do not change 
, your Wrangler configuration, or deployed Cron Triggers. Local development does not run cron schedules automatically.
Full logs and traces are available in the development session that owns the Worker when that session has 
The following example uses port 
. Replace it with your development server port, such as 
The Local Explorer API exposes the same operation for tools and AI agents. The request targets an exact Worker name and accepts a cron expression and optional scheduled time:
 "http://localhost:8787/cdn-cgi/local/explorer/api/local/scheduled?worker=my-worker"
 "Content-Type: application/json"
 '{"cron":"0 * * * *","scheduled_time":1745856238000}'
For more details, refer to 
Test 

Your builds now run on newer versions of Node.js, Python, Ruby, Go, Bun, Yarn, and Python package managers by default, so you get the latest fixes and features without changing your build…

AI Search keyword search supports Japanese, Chinese, and Korean queries
 tokenizer now matches Japanese, Chinese, and Korean queries, including full natural-language questions such as 
. Previously, a query without spaces was treated as a single exact phrase and usually returned no keyword results.
 tokenizer does not work well for these languages. If your content is in Japanese, Chinese, or Korean, set 
. Changing the tokenizer triggers a full reindex.
Search Cloudflare documentation with the Cloudflare CLI
, so you and your coding agent can find Cloudflare documentation without leaving the terminal. Search for guides, tutorials, and examples across Cloudflare products.
To install or update to the latest release:
 install --global cf@latest
Describe what you want to do, or ask a question:
 "rate limit requests with WAF"
The command searches live Cloudflare documentation and returns up to five ranked JSON results. Each result includes a title, description, and documentation URL. An Internet connection is required.
Agents can parse the results and fetch a complete page as Markdown. For example, use 
 to select the first result, then request its URL with the 
This release also standa

Cloudflare WAN and Magic Transit IPsec and GRE tunnels now support BGP import and export filtering.

Application Library now supports custom applications
Starting today, you can create 
 and reference them in Gateway DNS, Network, and HTTP policies. You can define an application with hostnames, IP subnets, or both. Then, optionally add TCP or UDP ports or port ranges to restrict HTTP and Network matching.
This release also improves search and filtering in the Application Library, updates the interfaces, data, and metrics used to review applications, and makes it easier to select and manage applications in the Gateway policy builder.
Applications and app types
Validate data loss prevention matches with Cloudflare Workers
Data loss prevention (DLP) custom pattern entries can use Cloudflare Workers in your account to validate matches with organization-specific logic. Your Worker can adjust a match's confidence or reject the match.
For setup steps and the Worker contract, refer to 
Validate data loss prevention matches with Cloudflare Workers
BGP route filtering for IPsec and GRE tunnels
Cloudflare WAN and Magic Transit IPsec and GRE tunnels now support BGP import and export filtering.
Create reusable filter profiles that specify an 
 action and a list of Classless Inter-Domain Routin

Accounts are no longer limited to 5 numbered sets. Create a set once and assign it to one or more zones in the same account.

Infrastructure applications support a configurable session duration
Infrastructure applications
 now have a configurable session duration with a default value of 24 hours. You can set the session duration when creating or editing an infrastructure application in the dashboard or via the API.
When the session duration elapses, any active SSH connections through that application will be closed, and users will need to reconnect. Previously, infrastructure applications did not enforce a session duration limit. Sessions could persist beyond 24 hours in some cases, but were never guaranteed beyond 10 hours due to 
Gateway connection limits
When a target matches multiple infrastructure applications, the shortest session duration applies.
Configure Advanced Custom Nameservers with nameserver sets (beta)
 are now available in beta to Enterprise accounts. A nameserver set groups two to five custom nameserver names with Cloudflare-assigned IP addresses. You can assign the set to zones in the same account.
Enterprise customers using Cloudflare DNS can create Advanced Custom Nameserver sets. These sets distribute addresses across Advanced anycast network groups to improve resiliency.
 sets, mod

Bot scores, verified bot status, and JA3/JA4 fingerprints no longer represent the internal connection between Cloudflare services.

Cloudflare API MCP server sends Workers Issues as MCP Events
 provides built-in error monitoring for Cloudflare Workers. It detects production failures and groups related failures into issues without an SDK or application wrapper.
 now sends these issues as 
, the first event it exposes. MCP clients discover 
cloudflare.alert.workers_observability_real_time_issue
, scoped to an account or a single Worker. When Issues detects a new or recurring issue, the server posts a signed event to the client's webhook, so an agent can start investigating without polling.
 on your Worker and connect 
 with the Notifications Write and Workers Observability Write permissions to a client that supports MCP Events webhooks.
New persistent browser tool in the Agents SDK, with tabs and logins that last between turns
 browser it can return to on later turns. Your agent can open a website, inspect a page, and come back to it with its tabs, cookies, and login session intact while the browser session remains alive.
 to your agent, the model writes code that drives the browser through the 
Chrome DevTools Protocol (CDP)
. Screenshots return as images that an image-capable model can inspect and your UI can d

As part of this change, the Zero Trust Network items were removed in favor of the new experiences available

Cloudflare One Client for Windows (version 2026.8.2110.0)
This hotfix resolves an issue where the client could repeatedly disconnect and reconnect when a split tunnel configuration with hostname-based entries failed to apply.
Cloudflare One Client for macOS (version 2026.8.2110.0)
This hotfix resolves an issue where the client could repeatedly disconnect and reconnect when a split tunnel configuration with hostname-based entries failed to apply.
Cloudflare One Client for Linux (version 2026.8.2110.0)
This hotfix resolves an issue where the client could repeatedly disconnect and reconnect when a split tunnel configuration with hostname-based entries failed to apply.
New web_fetch tool in the Agents SDK for reading webpages and documents, with automatic markdown conversion
 tool that reads webpages, PDFs, and supported Office documents, using 
Workers AI Markdown Conversion
 to turn them into text the model can read. The tool requests Markdown directly from sites that support 
 and formats JSON for the model.
 tool to your agent, the model provides a URL and receives back the page content, along with the final URL, HTTP status, and content type. Long pages come back in smaller sectio

You can now manually request an API Shield schema learning run without waiting for the weekly schedule. Use this during testing after sending representative traffic to your API.

Run schema learning on demand
You can now manually request an API Shield schema learning run without waiting for the weekly schedule. Use this during testing after sending representative traffic to your API.
 menu for the operations table and select 
. The run covers the entire zone and uses observed traffic. Existing learning thresholds still apply.
The dialog shows the learning status and last completed run for the zone. The 
 view shows the learning result and last learning run for each operation. You can review individual results while the zone-wide run continues.
For instructions and traffic requirements, refer to 
Run schema learning manually

is removed from converted responses rather than recalculated, because the Markdown body is streamed.

Cloudflare API MCP server serves Cloudflare skills
Cloudflare API MCP server
Skills over MCP extension
. MCP clients that support the extension discover the skills with 
https://mcp.cloudflare.com/mcp
MCP client that supports the extension
Improved HTTP/3 client cancellation reporting
Cloudflare has improved how it handles and reports client-cancelled HTTP/3 requests across Free, Pro, Business, and Enterprise plans. Customers now get a clearer view of client behavior in Cloudflare analytics and, where available, logs.
Previously, Cloudflare did not always stop an HTTP/3 request when the client cancelled its request stream. Some cancellations were already recorded as 
, while others continued to the origin and showed the eventual upstream status.
Cloudflare now stops affected requests sooner, reducing unnecessary origin work, and records them as 
. Customers may notice more 
 status codes for HTTP/3 traffic. This reflects more consistent reporting of existing cancellations, not an increase in failed requests.
 status codes in availability calculations should consider excluding them from server-side error rates because they represent requests cancelled by clients.
More efficient Mark

Existing WAF custom rules remain valid, but rules that reference a removed or renamed category will no longer match that category.

Create Workflow instance batches by count or list
 now accepts an options object that creates up to 100 Workflow instances in one call. The result lists the created instances and explains why any others were not created. To use this form in local development and get its types from 
, use Wrangler 4.148.0 or later.
To create instances that share the same options, pass 
. Each instance receives a generated ID:
To give each instance its own ID or options, pass 
(error.index, error.id, error.code, error.message);
(error.index, error.id, error.code, error.message);
 contains the created instances. 
 contains each entry that was not created, identified by its position in the input. IDs that already exist and IDs repeated within the batch are reported as errors instead of being skipped silently.
 is deprecated. Existing code that uses the array form continues to work.
Updated unsafe topic detection for AI Security for Apps
AI Security for Apps now supports an updated set of categories for detecting unsafe topics in incoming prompts.
cf.llm.prompt.unsafe_topic_categories
 have changed. Existing WAF custom rules remain valid, but rules that reference a removed or renamed category will no lo

Traffic to split tunnel excluded resources is no longer briefly blocked while the client is connecting or reconnecting.

Cloudflare One Client for macOS (version 2026.8.2100.0)
A new GA release for the macOS Cloudflare One Client is now available on the 
stable releases downloads page
This release includes the following highlights:
Traffic to split tunnel excluded resources is no longer briefly blocked while the client is connecting or reconnecting. The client now keeps its learned split tunnel configuration across reconnects.
Support for routing non-RFC 1918 local IPv4 networks through the tunnel when unrestricted LAN inclusion is enabled by policy or MDM.
Faster connects and lower memory use. The hosts file is now read once and shared across the client’s DNS resolvers instead of being reloaded by each one.
Additional changes and improvements
Improved reauthentication reliability and fixed an issue where a reauthentication could force a new registration.
Improved client reaction to the current network lowering its MTU.
Improved DNS reliability on networks with lower MTUs by clamping the TCP maximum segment size (MSS) for DNS-over-HTTPS connections sent through the tunnel.
Individual DNS-over-HTTPS queries now time out instead of hanging when the upstream server stops responding.
Improved API reliabi

This release introduces a new detection to mitigate a heap-based buffer overflow vulnerability in F5 BIG-IP, and enhances existing command injection protections by incorporating tested beta logic…

Standardize provider credential error responses in AI Gateway
AI Gateway's REST API now returns consistent responses when an AI provider rejects credentials. The change applies to 
Previous AI Gateway response
 for rejected credentials, with upstream retries
; the request fails without retrying the provider
 or another provider-specific status for rejected credentials
, the provider rejects the credentials
Provider-specific authentication error
Update applications that handle AI Gateway REST API errors to treat HTTP 
 as an invalid or rejected provider credential.
For details about providing provider credentials, refer to 
Bring your own provider keys
Warnings when approaching your DNS records quota
The DNS records page in the Cloudflare dashboard now shows a warning once you have used 85% of your 
The warning reflects the quota that applies to you. If your zone has its own quota, the warning shows that zone's usage. If your account uses an 
account-level DNS records quota
, the warning shows your usage across all zones in the account.
This release introduces a new detection to mitigate a heap-based buffer overflow vulnerability in F5 BIG-IP, and enhances existing command injection

BGP peering exchanges routes dynamically between your devices and your Cloudflare virtual network routing table. You no longer need to update static routes manually as your network changes.

BGP over IPsec and GRE tunnels generally available
BGP peering over IPsec and GRE tunnels is generally available for Cloudflare WAN and Magic Transit. You can use it for production workloads.
BGP peering exchanges routes dynamically between your devices and your Cloudflare virtual network routing table. You no longer need to update static routes manually as your network changes.
BGP over IPsec and GRE tunnels is available to all accounts that use 
. No enablement is required. BGP over CNI remains in closed beta.
For configuration details, refer to:
Configure BGP routes for Cloudflare WAN
Configure BGP routes for Magic Transit
 entry point supports AI Gateway and Workers AI models, so you can get started with Cloudflare models right away or use your existing Pi AI provider.

These databases run and persist data within the United States.

United States jurisdiction
You can create D1 databases with the 
 jurisdiction. These databases run and persist data within the United States.
Use this option for regional data residency requirements.
To create a database with the 

CLI, or the REST API, and cannot be added or changed afterwards.

Introducing Web Search API
 is now available in beta. Web Search API lets your AI agents and applications search the Internet and ground their responses in live information, instead of guessing URLs or relying on a model's training cutoff.
At launch, you can choose between three search providers: 
Ceramic.ai, Exa, and Linkup
. All three support Zero Data Retention for requests made through Cloudflare, and all have committed to Cloudflare's 
Web Search API runs through 
, so search requests appear in your gateway logs and are billed to your AI Gateway credits at each provider's list API price, with no additional markup. You can also bring your own provider API key.
Call Web Search API with the REST API:
 https://api.cloudflare.com/client/v4/accounts/
 "query": "What are some fun things to do in Salt Lake City as fall approaches?",
 "options": { "gateway": { "id": "default" } }
Or from a Worker with the AI binding:
"What are some fun things to do in Salt Lake City as fall approaches?"
How to use Web Search API
New strict service token authentication setting for Access
The strict service token authentication setting applies consistent behavior to requests made with service tokens. Whe

Workers AI embedding and reranking calls made by AI Search are included in AI Search pricing. These calls no longer appear on your Workers AI bill or in your AI Gateway logs.

Protect Quick Tunnels with email authentication
You can now restrict who can access a 
 to require visitors to authenticate with a one-time PIN sent to their email before they reach your local service.
Previously, anyone with a 
 URL could access the service behind it. Protected Quick Tunnels let you share a local development server, webhook receiver, or demo with specific people without creating a Cloudflare account or configuring a domain.
--allowed-mail alice@example.com
Multiple email addresses, by repeating the flag or using a comma-separated list: 
--allowed-mail 'alice@example.com,bob@example.com'
Every address on a domain: 
--allowed-mail '*@example.com'
Visitors do not need a Cloudflare account. Access ends for everyone when you stop the 
 to the latest version and refer to 
Introducing Clef: Cloudflare's first open-source decision models, now on Workers AI
@cf/cloudflare/clef-flash
, the first models trained by the Cloudflare Workers AI team, available on Workers AI today.
Clef is a decision model, in the same family as 
. Instead of generating text, it reads an input state and a set of typed questions, then returns a probability for every allowed answer. Your agent gets 

Existing Cloudflare Pipelines, R2 Data Catalog, and R2 SQL resources and configurations will continue to work and will be deprecated over time.

Basin Pipelines ingest limit increased to 1 GB/s
 can now ingest up to 1 GB/s, increased from 5 MB/s.
The higher per-stream limit gives high-volume application events, telemetry, and logs more room to grow without splitting ingestion across streams solely to stay within the previous limit.
For the full list of stream, sink, and pipeline limits, refer to 
Cloudflare Basin is now generally available
, formerly the Cloudflare Data Platform, is now generally available. Basin brings an end-to-end analytics platform to the Developer Platform, enabling you to collect data from a variety of sources, such as apps, infrastructure, devices, and other Cloudflare services, then query it to answer analytical questions.
, formerly Cloudflare Pipelines, ingests events from Workers, HTTP endpoints, and Cloudflare 
. It transforms events with SQL and ingests them into Iceberg tables or files on R2. With Basin Pipelines you can:
Ingest application and device events through HTTP endpoints or Workers bindings.
Filter and reshape Cloudflare logs before storing them as Iceberg tables, Parquet, or JSON.
Catch schema mismatches with typed bindings and investigate dropped events in the dashboard.
, formerly

Previously, purging by cache tag, hostname, prefix, or everything marked matching Cache Reserve content for revalidation. Purging by URL already removed content from Cache Reserve and is unchanged.

Connect multiple clients to one Browser Run session
 sessions now accept multiple concurrent connections. Before, a session accepted only one connection at a time, and other Workers had to wait until that connection closed. Now multiple Workers can connect to the same browser at the same time.
 call opens its own Chrome DevTools Protocol (CDP) connection. Create a separate browser context for each request to keep its pages, cookies, and storage apart from other clients.
// keep the shared browser running
// keep the shared browser running
Sharing sessions means fewer new browsers to launch, less cold-start time, and fewer 
 counted against your limits.
Concurrent connections require 
Browser Run adds WebMCP to Kitesurf and moves to document.modelContext
WebMCP Community Group draft
. Lab sessions no longer expose 
navigator.modelContextTesting
To list and run page tools:
 panel in the live view of a Lab session or in the 
--category-experimental-webmcp
Invalidate cached content instead of purging it
You can now invalidate cached content instead of purging it. Invalidation marks matching content as stale. On the next request, Cloudflare revalidates the content with your origin. If y

in Workers now support more of the OpenTelemetry span API, so you can instrument more of your code and record errors directly on your spans.

Subscribe to Browser Run crawl events
 can publish lifecycle events to 
. Subscribe to started, updated, and finished events to track progress or trigger downstream processing without polling.
To create an account-level subscription, run the following command:
 wrangler queues subscription create <QUEUE_NAME> --source browserRun --events crawl.started,crawl.updated,crawl.finished
 --source browserRun --events crawl.started,crawl.updated,crawl.finished" aria-label="Copy to clipboard" class="border-border text-muted-foreground hover:text-foreground focus-visible:outline-ring m-0 flex shrink-0 cursor-pointer items-center justify-center border-0 border-l border-solid bg-transparent px-3 transition-colors focus-visible:outline-2 focus-visible:-outline-offset-2">
 wrangler queues subscription create <QUEUE_NAME> --source browserRun --events crawl.started,crawl.updated,crawl.finished
 --source browserRun --events crawl.started,crawl.updated,crawl.finished" aria-label="Copy to clipboard" class="border-border text-muted-foreground hover:text-foreground focus-visible:outline-ring m-0 flex shrink-0 cursor-pointer items-center justify-center border-0 border-l border-solid bg-transparent px-3 t

Search devices, filter participants by type or status, open device details, and load additional results from each participant source.

Traffic Destination selector in Gateway policies
 selector that identifies how traffic exits Cloudflare. This allows administrators to write policies that target specific off-ramp methods - for example, applying different rules to traffic destined for the public Internet compared to traffic routed through Cloudflare Tunnel or Cloudflare WAN.
Available traffic destination values
Traffic to the public Internet
Traffic to a private origin through 
Traffic to another device running the 
 API field in both HTTP and Network policies.
net.offramp.type == "internet"
Add Mesh participants with guided onboarding
 now makes it faster to add and manage participants from the dashboard. Select 
 or find the information needed to connect a 
The updated dashboard includes the following improvements:
More Mesh node deployment options
 — Install a node on Linux, Kubernetes, Docker Compose, or Docker CLI. The dashboard provides requirements, commands, configuration, and links for each method. Refer to 
Run Mesh in Docker / Kubernetes
 for container deployment details.
Client device installation guidance
 — Access platform-specific 
 installers, mobile QR codes, and your Cloudflare One organization na

This release introduces new threat detections to enhance protection against Server-Side Request Forgery (SSRF) attempts using non-standard IP notations or jar loopback payloads, alongside new defenses…

Automatically manage inactive Access service tokens
Cloudflare Access administrators can now automatically disable or delete inactive service tokens. Administrators can set an inactivity period from 30 to 365 days and choose what Access does when a token reaches that limit.
To be eligible for cleanup, a token must be older than the configured period, must not have successfully authenticated during that period, and must not be directly referenced by an Access policy rule. Cleanup runs gradually in the background, so eligible tokens may not be disabled or deleted immediately.
For configuration instructions, refer to 
Manage inactive service tokens
Private MCP server support for MCP server portals
 can now connect to MCP servers available only on your private network. The portal uses 
 and IP addresses without exposing the MCP server to the public Internet.
Connect the server network to Cloudflare with 
. Configure a private hostname or CIDR route, then turn on 
Route traffic through Cloudflare Gateway
 when you add the server. OAuth authorization server endpoints, such as the authorization and token endpoints, must be accessible on the public Internet. If Cloudflare automatically reg

Windows 10, the last Windows release to support 32-bit systems, reached end of support in October 2025. Apple has also deprecated Intel-based Mac computers.

Inspect logs, network requests, and DOM in Session Recordings
Browser Run Session Recordings
 panel, giving you more context to understand what happened during a browser session without having to reproduce it.
 tab lets you search captured console output and filter messages by level. The 
 tab shows each request's method, status, headers, payload, response, and timing waterfall, with the option to download the session's network activity as a HAR file.
retrieve recorded network activity via API
 as raw JSON or a HAR file for use in your own debugging and analysis workflows.
 tab provides an expandable view of the page structure at the end of the recording and lets you copy the reconstructed HTML. For sessions with multiple browser tabs, the Inspect panel updates to show data for the tab selected in the recording viewer.
To get started, enable recording when launching a browser session. After the session closes, open 
 and select the recording icon next to the session.
Session recording documentation
 for setup instructions and current limits.
Unified Routing generally available
Unified Routing is generally available for Cloudflare WAN and Magic Transit.
Unified Routing improves the 

Workers traces can now follow JavaScript RPC calls across Worker boundaries and into Durable Objects. Previously, a trace stopped at the caller's RPC boundary.

Delete Workflow instances individually or in batches
You can now delete one or up to 100 Workflow instances and their stored state via the 
 or Wrangler 4.125.0 and later. Deleting an instance frees its stored state and stops its current execution. 
 is based on the average daily peak.
Delete one instance by calling 
If a Workflow deletes its own instance, execution stops during 
. Code after the call does not run.
Delete multiple instances by calling 
The batch result contains 
 entries for successful deletions and per-instance errors. IDs that do not exist are returned as errors. Duplicate IDs count toward the limit and are deleted once, with the result repeated for each input position.
Wrangler accepts positional instance IDs, a file containing a top-level JSON array of strings, or both, up to 100 IDs total. Use 
 to delete the most recently created instance. Use 
Delete Workflow instances
Create additional Free accounts through the dashboard and API
We're expanding how customers create accounts across Cloudflare, making it easier to self-serve account creation in the dashboard, automate standalone account creation with user-owned API tokens or OAuth access tokens, and create Fr

Cloudflare now displays warnings for shadowed records in all zones. A record is shadowed when a subdomain delegation gives authority for its name, or a name below it, to another set of nameservers.

Hyperdrive support for Python Workers
 can now connect to PostgreSQL and MySQL through Hyperdrive.
For setup, code examples, and limitations, refer to 
Use Hyperdrive from Python Workers
Stream Workflow instance events in your Worker or via the API with .subscribe()
You can now stream Workflow instance events via 
WorkflowInstance.subscribe()
 API endpoint. Workers and HTTP clients can react to 
 events, including attempts, sleeps, waits, and rollbacks, without polling for instance status.
A subscription first streams the entire event history of the Workflow instance. After streaming past events, the subscription waits for new events as the instance runs. You can use 
 to receive only specific event types or 
 to start a subscription at a specific event.
 to update Workflow status in user-facing dashboards, send notifications when steps complete, or trigger follow-up work for specific events.
For event types, available fields, and subscription options, refer to 
Access for Infrastructure now supports tagged targets and tag-based target criteria
Access for Infrastructure
. You can attach key-value tags to 
 and use them in access policies.
You can manage tags on targets inline when 

The Rulesets API now supports Origin Range Requests in Cache Rules. This setting lets Cloudflare fetch large files from your origin in cache-aligned byte ranges.

Require fresh authentication for SAML identity providers
Cloudflare Access can now request fresh authentication from a SAML identity provider for every login. Turn on 
 in the Cloudflare dashboard, or set 
 through the API. Access will then set 
 in signed and unsigned SAML authentication requests.
This option is useful when an application requires users to reauthenticate at the identity provider instead of relying on an existing identity provider session. The default value is 
For configuration details, refer to 
Require fresh authentication at the identity provider
Control which hostnames Browser Run sessions can access
, which limit a browser session's HTTP and HTTPS requests to permitted hostnames.
Use guardrails when you need to:
Keep a browser workflow limited to a specific website and its subdomains.
Load only known third-party APIs, scripts, images, and fonts.
Generate a screenshot or PDF from HTML you provide while preventing it from loading external content.
Set guardrails when starting a session with Puppeteer, Playwright, or the REST API. With a browser binding named 
 when launching Puppeteer:
In addition to session guardrails, Browser Run now supports a read-only mode

The browser console combines server lifecycle events with local microphone, connection, and playback events, including model start, first model text, first audio, and playback start.

Inspect Voice Agent turn latency and outcomes
 v0.4.0 now lets you inspect where each Voice Agent turn spends time and how it ends.
(turn.outcome, turn.turnTotalMs);
 package lets you build real-time voice agents with Cloudflare Agents. It streams microphone audio to an Agent over WebSocket, transcribes speech, runs your model through 
, converts the response to speech, and streams audio back to the caller.
A turn moves through several stages:
User speaks -> speech-to-text -> model -> text-to-speech -> audio
Previously, the package's four aggregate metrics covered successful, non-empty speech turns. They did not show how failed, aborted, empty, or text turns ended.
Each speech or text turn now produces a typed 
 for correlating events from the same turn.
A terminal outcome such as 
Timings for important stages, including speech-to-final-transcript, model-to-first-text, TTS-to-first-audio, and total turn duration.
These timings can overlap and are not additive. Timings for stages that a turn did not reach are omitted.
The latest summary is available through 
. Voice input includes only the speech and transcription timings it can measure.
If an agent produces no audio, you can now di

The client no longer requires the Windows WLAN AutoConfig service to be running.

Default instance retention for new Workflows on Workers Paid is seven days
 created on or after September 10, 2026, on the Workers Paid plan retain completed and errored instance state for seven days by default (previously 30 days). The seven day default helps to reduce storage costs by default. The maximum retention 
The retention period for existing Workflows is unchanged. The Workers Free plan retains its three-day default and limit.
To set the retention period for a Workflow instance, specify 
You can also set the retention period per Workflow and per instance in the 
For retention details, refer to 
WorkflowInstanceCreateOptions
Use Cloudflare Containers with Codex via the OpenAI Agents API
The OpenAI Agents API gives your application access to Codex through an OpenAI-managed API.
OpenAI manages sessions, orchestration, context compaction, and recovery while your application provides tools and uses Cloudflare Containers as the execution environment.
Cloudflare Containers can now provide self-hosted execution environments for the OpenAI Agents API. The open-source 
OpenAI Agents API Workers template
 provides a reference implementation. The Worker maintains a Cloudflare Contain

AI Gateway custom costs now support cache-read and cache-write token rates. This lets custom cost metrics reflect negotiated cache pricing across providers.

AI Gateway custom costs support cache tokens
AI Gateway custom costs now support cache-read and cache-write token rates. This lets custom cost metrics reflect negotiated cache pricing across providers.
Cache-token pricing activates when either cache rate is present. An omitted cache rate defaults to 
. If both cache rates are omitted, AI Gateway preserves the existing input and output calculation.
Providers can include cache tokens within input tokens or report them separately. AI Gateway automatically accounts for these differences and prevents double-counting.
Improved iOS tap-to-type experience for Browser Isolation
 has improved the tap-to-type experience for users on iOS devices.
Previously, Browser Isolation displayed a full-screen overlay with the message 
 when users focused a text field. The prompt now appears inline over the focused text field, reducing disruption when users enter text in isolated sessions.
If the focused text field is too small to display the full prompt, Browser Isolation displays a keyboard icon in the center of the text field instead.
iOS users should tap twice to begin entering text. This update applies automatically to Browser Isolation sessions on 

Removed Miniflare's built-in module discovery; higher-level tools like Wrangler and the Vite plugin should be providing the module graph.

Radar search now includes Internet events
 search now includes Internet events and outages alongside existing results. Search event descriptions or related entities, such as locations, ASes, bots, and top-level domains, to find relevant events and open the most relevant Radar view.
Event links preserve the event date range, making it easier to investigate what changed before, during, and after an event. These results are also available to browser-based AI agents through 
This release enhances detection logic for existing rules targeting Next.js remote code execution (RCE) vulnerabilities by consolidating active beta rules into baseline signatures.
This rule is merged into the original rule "Next.js - Image Optimizer Remote Code Execution via Crafted AVIF" (ID: 
This rule is merged into the original rule "Next.js - Remote Code Execution - CVE:CVE-2026-75604" (ID: 
WAF Release - Scheduled changes for 2026-09-15
Version Control - Information Disclosure - Beta
This rule will be merged into the original rule "Version Control - Information Disclosure" (ID: 
Command Injection - Generic 10
Miniflare v5 prepares local development for the cf CLI
Miniflare v5 prepares Cloudflare local develop