# \[ngrok news\]

ngrok is a globally distributed reverse proxy that secures, protects and accelerates your applications and network services, no matter where you run them.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## Coding agent, meet ngrok: the ngrok MCP server is live

DevFeed: [Coding agent, meet ngrok: the ngrok MCP server is live](<https://devfeed.tech/articles/coding-agent-meet-ngrok-the-ngrok-mcp-server-is-live-136004.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/mcp>)

Author: Joel Hans

Published: 2026-11-16T00:00:00Z

Content type: release

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [cursor](<https://devfeed.tech/topics/cursor.md>), [codex](<https://devfeed.tech/topics/codex.md>), [Claude Code](<https://devfeed.tech/topics/claude-code.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [2026](<https://devfeed.tech/tags/2026.md>), [3](<https://devfeed.tech/tags/3.md>), [404](<https://devfeed.tech/tags/404.md>), [about](<https://devfeed.tech/tags/about.md>), [account](<https://devfeed.tech/tags/account.md>), [agent](<https://devfeed.tech/tags/agent.md>), [ai](<https://devfeed.tech/tags/ai.md>), [code](<https://devfeed.tech/tags/code.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agent](<https://devfeed.tech/tags/coding-agent.md>), [free](<https://devfeed.tech/tags/free.md>), [internet](<https://devfeed.tech/tags/internet.md>), [live](<https://devfeed.tech/tags/live.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-server](<https://devfeed.tech/tags/mcp-server.md>), [needs](<https://devfeed.tech/tags/needs.md>), [ngrok](<https://devfeed.tech/tags/ngrok.md>), [plan](<https://devfeed.tech/tags/plan.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>), [server](<https://devfeed.tech/tags/server.md>), [traffic-policy](<https://devfeed.tech/tags/traffic-policy.md>), [works](<https://devfeed.tech/tags/works.md>)

### AI overview

ngrok's MCP server lets compatible coding agents inspect traffic and account resources, search current documentation, and create or update cloud endpoint Traffic Policy rules. In the gpu.compare example, the agent removed an OAuth gate, analyzed traffic, added scanner blocks and a per-IP rate limit, and configured a country cookie; it validated policy rules and checked their effects. The server provides tool descriptions, documentation search, and a policy validator to reduce configuration guesswork. Access is scope-based: users approve read or write permissions, and OAuth tokens are limited to the MCP server, expire after 15 minutes with refresh during work, and require renewed approval after a week. The server does not accept API keys or authtokens; a personal access token is an option for clients without OAuth. The MCP manages account-side configuration, while the ngrok CLI still creates tunnels. Availability and supported clients vary by the listed setup options; ChatGPT is stated not to connect yet.

### Source excerpt

The ngrok MCP server is live. Say what needs to happen between your code and the internet, and your coding agent turns it into ngrok. Works on every plan, even free.

## Serving models at scale with vLLM

DevFeed: [Serving models at scale with vLLM](<https://devfeed.tech/articles/serving-models-at-scale-with-vllm-121142.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/vllm>)

Author: Annie Sexton

Published: 2026-11-05T00:00:00Z

Content type: article

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [vllm](<https://devfeed.tech/topics/vllm.md>), [LLM optimization](<https://devfeed.tech/topics/llm-optimization.md>), [LLM Techniques](<https://devfeed.tech/topics/llm-techniques.md>), [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [Mixture of Experts (MoE)](<https://devfeed.tech/topics/mixture-of-experts-moe.md>), [quantization](<https://devfeed.tech/topics/quantization.md>), [sglang](<https://devfeed.tech/topics/sglang.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [2023](<https://devfeed.tech/tags/2023.md>), [ai](<https://devfeed.tech/tags/ai.md>), [batching](<https://devfeed.tech/tags/batching.md>), [caching](<https://devfeed.tech/tags/caching.md>), [continuous](<https://devfeed.tech/tags/continuous.md>), [continuous-batching](<https://devfeed.tech/tags/continuous-batching.md>), [gpus](<https://devfeed.tech/tags/gpus.md>), [inference](<https://devfeed.tech/tags/inference.md>), [keep](<https://devfeed.tech/tags/keep.md>), [learn](<https://devfeed.tech/tags/learn.md>), [llms](<https://devfeed.tech/tags/llms.md>), [models](<https://devfeed.tech/tags/models.md>), [more](<https://devfeed.tech/tags/more.md>), [prefix-caching](<https://devfeed.tech/tags/prefix-caching.md>), [scale](<https://devfeed.tech/tags/scale.md>), [serving](<https://devfeed.tech/tags/serving.md>), [uses](<https://devfeed.tech/tags/uses.md>), [vllm](<https://devfeed.tech/tags/vllm.md>), [with](<https://devfeed.tech/tags/with.md>), [works](<https://devfeed.tech/tags/works.md>)

### AI overview

vLLM improves open-weight LLM serving by reducing wasted GPU compute and KV-cache memory. Continuous batching adds waiting requests as soon as capacity opens; PagedAttention allocates KV data in fixed-size blocks as needed, and prefix caching reuses matching token prefixes. Its scheduler prioritizes ongoing decoding and fits prompt prefill into the remaining token budget; chunked prefill lets long prompts progress across passes, trading a longer wait for the first token for smoother ongoing responses. CUDA graphs reduce CPU launch gaps, while speculative decoding can generate several accepted tokens per expensive pass when spare compute is available, but may help less or slow service under heavy load. Quantization and tensor, pipeline, expert, or data parallelism address model size and traffic needs. The article says continuous batching, PagedAttention, prefix caching, chunked prefill, and CUDA graphs are enabled by default; speculative decoding, quantization, and parallelism require configuration. It describes SGLang as overlapping with vLLM and excelling on shared-prefix workloads, while vLLM supports more models and hardware and has a larger community.

### Source excerpt

Serving LLMs at scale is hard. Learn how inference works and how vLLM uses continuous batching, PagedAttention, prefix caching, and more to keep GPUs busy.

## What the heck is the Public Suffix List?

DevFeed: [What the heck is the Public Suffix List?](<https://devfeed.tech/articles/what-the-heck-is-the-public-suffix-list-96025.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/public-suffix-list>)

Author: Joel Hans

Published: 2026-10-22T00:00:00Z

Content type: article

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [trust](<https://devfeed.tech/topics/trust.md>), [hosting](<https://devfeed.tech/topics/hosting.md>), [Firefox](<https://devfeed.tech/topics/firefox.md>), [rate-limiting](<https://devfeed.tech/topics/rate-limiting.md>)

Tags: [account](<https://devfeed.tech/tags/account.md>), [administrator](<https://devfeed.tech/tags/administrator.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [amazon](<https://devfeed.tech/tags/amazon.md>), [app](<https://devfeed.tech/tags/app.md>), [app-updates](<https://devfeed.tech/tags/app-updates.md>), [applications](<https://devfeed.tech/tags/applications.md>), [apps](<https://devfeed.tech/tags/apps.md>), [automate](<https://devfeed.tech/tags/automate.md>), [file](<https://devfeed.tech/tags/file.md>), [hosting](<https://devfeed.tech/tags/hosting.md>), [list](<https://devfeed.tech/tags/list.md>), [platforms](<https://devfeed.tech/tags/platforms.md>), [public](<https://devfeed.tech/tags/public.md>), [sites](<https://devfeed.tech/tags/sites.md>), [text](<https://devfeed.tech/tags/text.md>), [unrelated](<https://devfeed.tech/tags/unrelated.md>)

### AI overview

The Public Suffix List (PSL) gives software a shared map of domain boundaries that DNS does not reliably provide for determining which sites belong to separate users. Browsers use it to reject cookies scoped across unrelated tenants; Firefox uses it to emphasize registrable domains, and password managers and certificate authorities use it for matching and rate-limit grouping. Platforms also use it for account and billing tasks. Replacing the list has been difficult: DNS-based proposals faced disagreement over boundary needs and concerns about extra lookups, though DMARCbis adopted a DNS tree walk for email policy. The private section has grown, increasing review work for volunteer maintainers. Proposed entry checks include evidence of independent users, domain administrator control, and sufficient registration time. Maintainers disagree over whether to tighten or relax the current user threshold, and accepted changes reach downstream software only when its maintainers adopt them. The article also notes that rate limits based on PSL entries can be gamed and are discouraged as a use of the list.

### Source excerpt

How a volunteer-maintained text file helps keep unrelated sites apart, why it's hard to replace, and what hosting platforms ask of the people maintaining it.

## The eMudhra CA compromise and ngrok: what's affected and what to do

DevFeed: [The eMudhra CA compromise and ngrok: what's affected and what to do](<https://devfeed.tech/articles/the-emudhra-ca-compromise-and-ngrok-what-s-affected-and-what-to-do-93899.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/emudhra-ca-compromise>)

Author: Sam Richard

Published: 2026-10-21T00:00:00Z

Content type: news

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Exploit](<https://devfeed.tech/topics/exploit.md>), [enterprise certificate authority certificates](<https://devfeed.tech/topics/enterprise-certificate-authority-certificates.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [api](<https://devfeed.tech/tags/api.md>), [api-key](<https://devfeed.tech/tags/api-key.md>), [ca](<https://devfeed.tech/tags/ca.md>), [cas](<https://devfeed.tech/tags/cas.md>), [certificate](<https://devfeed.tech/tags/certificate.md>), [certificates](<https://devfeed.tech/tags/certificates.md>), [compromise](<https://devfeed.tech/tags/compromise.md>), [config](<https://devfeed.tech/tags/config.md>), [configuration](<https://devfeed.tech/tags/configuration.md>), [connect](<https://devfeed.tech/tags/connect.md>), [emudhra](<https://devfeed.tech/tags/emudhra.md>), [emudhra-ca-compromise](<https://devfeed.tech/tags/emudhra-ca-compromise.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [go](<https://devfeed.tech/tags/go.md>), [https](<https://devfeed.tech/tags/https.md>), [ngrok](<https://devfeed.tech/tags/ngrok.md>), [ngrok-agent](<https://devfeed.tech/tags/ngrok-agent.md>), [operating-system](<https://devfeed.tech/tags/operating-system.md>), [os](<https://devfeed.tech/tags/os.md>), [security](<https://devfeed.tech/tags/security.md>)

### AI overview

The eMudhra compromise exposed private keys for eight intermediate CAs, which could let an attacker with a privileged network position intercept TLS connections using certificates chained to affected roots. Default ngrok agent tunnels are not affected: the agent pins ngrok's root CA, and signed updates are verified independently of TLS. Host trust-store use may create exposure for `ngrok api`, `--traffic-policy-url`, HTTPS `proxy_url`, or `connect_cas: host`. Check the config with `ngrok config check` and search scripts and service definitions for those features. If applicable, update the host CA bundle to remove affected roots and restart the agent, or upgrade to v3.39.12, which uses a patched Go release that blocklists them. The excerpt says Windows users should use the direct download because the Microsoft Store build may take days to update; agent v2 is past end-of-life.

### Source excerpt

If you run the ngrok agent with its default configuration, your tunnels aren't affected by the eMudhra CA compromise. Here's which settings are, and what to do.

## Add ngrok to Next.js, FastAPI, Axum, and Gin

DevFeed: [Add ngrok to Next.js, FastAPI, Axum, and Gin](<https://devfeed.tech/articles/add-ngrok-to-next-js-fastapi-axum-and-gin-90311.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/ngrok-web-framework-packages>)

Author: Ishan Jain

Published: 2026-10-19T00:00:00Z

Content type: tutorial

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [FastAPI](<https://devfeed.tech/topics/fastapi.md>), [Gin](<https://devfeed.tech/topics/gin.md>)

Tags: [fastapi](<https://devfeed.tech/tags/fastapi.md>), [gin](<https://devfeed.tech/tags/gin.md>), [next-js](<https://devfeed.tech/tags/next-js.md>), [ngrok](<https://devfeed.tech/tags/ngrok.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>), [sdks](<https://devfeed.tech/tags/sdks.md>)

### AI overview

Official ngrok packages for Next.js, FastAPI, Axum, and Gin open an app's public endpoint when its server starts and close it when the server stops. The article describes configuration options and ways to install and run each package.

### Source excerpt

Official ngrok packages for Next.js, FastAPI, Axum, and Gin open your app's public endpoint when your server starts and close it when it stops. No second terminal.

## How Kubernetes probes work

DevFeed: [How Kubernetes probes work](<https://devfeed.tech/articles/how-kubernetes-probes-work-83312.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/probes>)

Author: Sam Rose

Published: 2026-08-19T00:00:00Z

Content type: tutorial

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Amazon Elastic Kubernetes Service](<https://devfeed.tech/topics/amazon-elastic-kubernetes-service.md>), [k3s](<https://devfeed.tech/topics/k3s.md>)

Tags: [clusters](<https://devfeed.tech/tags/clusters.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [kubernetes-clusters](<https://devfeed.tech/tags/kubernetes-clusters.md>), [learn](<https://devfeed.tech/tags/learn.md>), [work](<https://devfeed.tech/tags/work.md>)

### AI overview

An interactive tutorial explains Kubernetes startup, readiness, and liveness probes, including configuration, failure modes, graceful pod termination, and effects on deployment rollouts. It also describes a reported Kubernetes bug in which liveness probes may run before a startup probe succeeds, and discusses how poorly chosen liveness checks and client retries can contribute to cascading failures.

### Source excerpt

Learn how probes work interactively on simulated Kubernetes clusters running in your browser.

## Compression is prediction

DevFeed: [Compression is prediction](<https://devfeed.tech/articles/compression-is-prediction-83205.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/compression-is-prediction>)

Author: Annie Sexton

Published: 2026-08-11T00:00:00Z

Content type: article

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Compression](<https://devfeed.tech/topics/compression.md>), [model architecture](<https://devfeed.tech/topics/model-architecture.md>), [Icon font](<https://devfeed.tech/topics/icon-font.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [arithmetic-coding](<https://devfeed.tech/tags/arithmetic-coding.md>), [code](<https://devfeed.tech/tags/code.md>), [compression](<https://devfeed.tech/tags/compression.md>), [data](<https://devfeed.tech/tags/data.md>), [decoding](<https://devfeed.tech/tags/decoding.md>), [encoding](<https://devfeed.tech/tags/encoding.md>), [llms](<https://devfeed.tech/tags/llms.md>)

### AI overview

The article explains how data compression uses models and entropy coding, including arithmetic coding, to represent data with fewer bits. It connects compression to large language models, which predict the next token from context, and explains that their model size and compute demands make them impractical for compressing routine web responses.

### Source excerpt

Compression and LLMs are trying to solve the exact same problem: predicting what comes next. Learn the fundamentals of compression and how better prediction leads to better shrinkage.

## Access a Raspberry Pi remotely over SSH with ngrok

DevFeed: [Access a Raspberry Pi remotely over SSH with ngrok](<https://devfeed.tech/articles/how-to-access-a-raspberry-pi-remotely-without-pesky-port-forwarding-83166.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/access-raspberry-pi-remotely>)

Author: Sam Richard

Published: 2026-08-04T00:00:00Z

Content type: tutorial

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [ssh](<https://devfeed.tech/topics/ssh.md>), [Raspberry Pi](<https://devfeed.tech/topics/raspberry-pi.md>), [Virtual Private Network](<https://devfeed.tech/topics/vpn.md>)

Tags: [cgnat](<https://devfeed.tech/tags/cgnat.md>), [firewalls](<https://devfeed.tech/tags/firewalls.md>), [https](<https://devfeed.tech/tags/https.md>), [nat](<https://devfeed.tech/tags/nat.md>), [networking](<https://devfeed.tech/tags/networking.md>), [ngrok](<https://devfeed.tech/tags/ngrok.md>), [ngrok-agent](<https://devfeed.tech/tags/ngrok-agent.md>), [raspberry-pi](<https://devfeed.tech/tags/raspberry-pi.md>), [secure-tunnels](<https://devfeed.tech/tags/secure-tunnels.md>), [ssh](<https://devfeed.tech/tags/ssh.md>), [vpn](<https://devfeed.tech/tags/vpn.md>)

### AI overview

A step-by-step guide to reaching a Raspberry Pi over SSH using an outbound ngrok tunnel, including when the device is behind NAT, CGNAT, or a firewall. It covers installing and configuring the agent, reserving a stable TCP address, running the tunnel as a background service, and restricting access with SSH keys, IP rules, and mutual TLS.

### Source excerpt

SSH into a Raspberry Pi from anywhere without port forwarding, a static IP, or a VPN. A step-by-step guide with ngrok that works behind NAT, CGNAT, and firewalls.

## ngrok's new Terraform provider

DevFeed: [ngrok's new Terraform provider](<https://devfeed.tech/articles/ngrok-s-new-terraform-provider-83339.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/terraform-provider>)

Author: Ishan Jain

Published: 2026-07-21T00:00:00Z

Content type: release

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Terraform](<https://devfeed.tech/topics/terraform.md>), [Secrets Management](<https://devfeed.tech/topics/secrets-management.md>)

Tags: [api](<https://devfeed.tech/tags/api.md>), [authentication](<https://devfeed.tech/tags/authentication.md>), [declarative](<https://devfeed.tech/tags/declarative.md>), [git](<https://devfeed.tech/tags/git.md>), [infrastructure](<https://devfeed.tech/tags/infrastructure.md>), [jwt](<https://devfeed.tech/tags/jwt.md>), [ngrok](<https://devfeed.tech/tags/ngrok.md>), [oauth](<https://devfeed.tech/tags/oauth.md>), [policy](<https://devfeed.tech/tags/policy.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>), [secrets](<https://devfeed.tech/tags/secrets.md>), [terraform](<https://devfeed.tech/tags/terraform.md>), [terraform-provider](<https://devfeed.tech/tags/terraform-provider.md>), [traffic-policy](<https://devfeed.tech/tags/traffic-policy.md>)

### AI overview

The rebuilt ngrok Terraform provider is available as v0.8.1. It lets teams manage domains, cloud endpoints, traffic policies, and credentials as code, with Terraform import support for existing resources.

### Source excerpt

The rebuilt ngrok Terraform provider is live as v0.8.1. Manage domains, cloud endpoints, Traffic Policy, and secrets as code, and import everything you're already running.

## Migrate to ngrok from Cloudflare Tunnels

DevFeed: [Migrate to ngrok from Cloudflare Tunnels](<https://devfeed.tech/articles/migrate-to-ngrok-from-cloudflare-tunnels-83272.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/migration-guide-alternative-to-cloudflare-tunnels>)

Author: Scott McAllister

Published: 2026-07-17T00:00:00Z

Content type: tutorial

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Developer Platform](<https://devfeed.tech/topics/developer-platform.md>), [Cloudflare](<https://devfeed.tech/topics/cloudflare.md>), [Network Services](<https://devfeed.tech/topics/network-services.md>), [Cloudflare Workers](<https://devfeed.tech/topics/cloudflare-workers.md>)

Tags: [cli](<https://devfeed.tech/tags/cli.md>), [cloudflare](<https://devfeed.tech/tags/cloudflare.md>), [docker](<https://devfeed.tech/tags/docker.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [migrate](<https://devfeed.tech/tags/migrate.md>), [ngrok](<https://devfeed.tech/tags/ngrok.md>), [secure-tunnels](<https://devfeed.tech/tags/secure-tunnels.md>)

### AI overview

A migration guide for software teams moving ingress from Cloudflare Tunnels to ngrok. It explains how to map tunnel ingress rules to ngrok endpoints and configure them through the CLI or agent configuration file, with options for SDKs, Docker, and Kubernetes. It also describes ngrok Traffic Policy features for authentication, filtering, rate limiting, and request transformation.

### Source excerpt

Migrate from Cloudflare Tunnels to ngrok in minutes. Map your cloudflared config to ngrok endpoints, add auth and routing with Traffic Policy, and run it anywhere -- CLI, SDKs, Docker, or Kubernetes.

## Traffic Policy is now 10x cheaper, and idle custom domains are free

DevFeed: [Traffic Policy is now 10x cheaper, and idle custom domains are free](<https://devfeed.tech/articles/traffic-policy-is-now-10x-cheaper-and-idle-custom-domains-are-free-83311.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/pricing-june-2026>)

Author: Sam Richard

Published: 2026-07-13T00:00:00Z

Content type: release

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [policy-as-code](<https://devfeed.tech/topics/policy-as-code.md>), [Load Balancing](<https://devfeed.tech/topics/load-balancing.md>)

Tags: [company](<https://devfeed.tech/tags/company.md>), [infrastructure](<https://devfeed.tech/tags/infrastructure.md>), [pricing](<https://devfeed.tech/tags/pricing.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>), [traffic-policy](<https://devfeed.tech/tags/traffic-policy.md>)

### AI overview

ngrok is lowering Traffic Processing Unit prices by 10x, billing custom domains only during hours when they receive traffic, and charging separately for each active endpoint in a pool. The pricing changes take effect on June 22 and June 29, respectively; most customers should pay less, while a small number of heavy Endpoint Pooling users may pay more.

### Source excerpt

Traffic Policy pricing drops by 10x, custom domains move to active-only billing, and Endpoint Pooling now bills per active endpoint. Here's what changed and why.

## The new ngrok.ai

DevFeed: [The new ngrok.ai](<https://devfeed.tech/articles/the-new-ngrok-ai-83281.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/new-ngrok-ai>)

Author: Nijiko Yonskai

Published: 2026-07-01T00:00:00Z

Content type: release

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [API Monetization](<https://devfeed.tech/topics/api-monetization.md>)

Tags: [access-controls](<https://devfeed.tech/tags/access-controls.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-gateway](<https://devfeed.tech/tags/ai-gateway.md>), [costs](<https://devfeed.tech/tags/costs.md>), [ngrok](<https://devfeed.tech/tags/ngrok.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>)

### AI overview

ngrok has moved its AI Gateway to app.ngrok.ai, with a dedicated API and gateway URL for applications using hosted or self-managed models. The gateway centralizes model routing, access controls, observability, and usage visibility. The announcement also describes a per-request processing fee and the wind-down of the old dashboard.

### Source excerpt

The ngrok AI Gateway is now app.ngrok.ai. One key, one URL to access any model, including the ones you run yourself, with access controls and per-call cost visibility.

## I ported Kubernetes to the browser

DevFeed: [I ported Kubernetes to the browser](<https://devfeed.tech/articles/i-ported-kubernetes-to-the-browser-83249.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/i-ported-kubernetes-to-the-browser>)

Author: Sam Rose

Published: 2026-06-30T00:00:00Z

Content type: article

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Test automation](<https://devfeed.tech/topics/test-automation.md>), [aosp](<https://devfeed.tech/topics/aosp.md>), [ai slop pull requests](<https://devfeed.tech/topics/ai-slop-pull-requests.md>)

Tags: [agents](<https://devfeed.tech/tags/agents.md>), [clusters](<https://devfeed.tech/tags/clusters.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [slop](<https://devfeed.tech/tags/slop.md>)

### AI overview

The author describes webernetes, a partial TypeScript port of Kubernetes that runs simulated clusters in a browser for interactive teaching content. They explain how they used LLMs to port much of the code, then reviewed it and built extensive unit and integration tests comparing its behavior with real Kubernetes running on k3s. The article discusses common LLM porting errors and argues that code review and tests are both necessary for confidence in generated code.

### Source excerpt

Almost 100,000 lines of LLM-generated code in 2 months, and none of it is slop.

## AOL was down (1996)

DevFeed: [AOL was down (1996)](<https://devfeed.tech/articles/aol-was-down-1996-83179.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/aol-was-down-1996>)

Author: Mac Chaffee

Published: 2026-06-23T00:00:00Z

Content type: opinion

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Post Mortem](<https://devfeed.tech/topics/post-mortem.md>), [engineering-leadership](<https://devfeed.tech/topics/engineering-leadership.md>)

Tags: [engineering](<https://devfeed.tech/tags/engineering.md>)

### AI overview

The author revisits AOL's 19-hour 1996 outage to argue that reliability postmortems should center the people affected, not only technical metrics. The essay connects outage impacts to economic incentives and unequal switching power, and suggests gathering customer impact accounts to support investment in reliability. It calls on SREs to coordinate standards and resist declining reliability and quality.

### Source excerpt

A historical journey through a 1996 AOL outage that captured America's full attention. Thirty years later, in a world where slop sells, what can SREs learn from that event?

## AI Gateway: attach provider keys to AI Gateway keys

DevFeed: [AI Gateway: attach provider keys to AI Gateway keys](<https://devfeed.tech/articles/ai-gateway-attach-provider-keys-to-ai-gateway-keys-83174.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/ai-gateway-provider-keys>)

Author: Nijiko Yonskai

Published: 2026-04-22T00:00:00Z

Content type: release

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Secrets Management](<https://devfeed.tech/topics/secrets-management.md>), [payment-gateway](<https://devfeed.tech/topics/payment-gateway.md>), [token jacking](<https://devfeed.tech/topics/token-jacking.md>)

Tags: [ai-gateway](<https://devfeed.tech/tags/ai-gateway.md>), [api-keys](<https://devfeed.tech/tags/api-keys.md>), [auth](<https://devfeed.tech/tags/auth.md>), [billing](<https://devfeed.tech/tags/billing.md>), [bring-your-own-key](<https://devfeed.tech/tags/bring-your-own-key.md>), [failover](<https://devfeed.tech/tags/failover.md>), [keys](<https://devfeed.tech/tags/keys.md>), [ollama](<https://devfeed.tech/tags/ollama.md>), [openai](<https://devfeed.tech/tags/openai.md>), [policy](<https://devfeed.tech/tags/policy.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>)

### AI overview

ngrok AI Gateway now lets users attach OpenAI, Anthropic, and other managed provider keys directly to AI Gateway keys. Applications authenticate with the gateway key, while the gateway uses the attached provider credential upstream. The change removes the need to keep provider keys in application code or Traffic Policy for managed providers, and includes encrypted storage, redaction after creation, and support for multiple keys for failover or staged rollouts. Existing Traffic Policy BYOK configurations continue to work but cannot be used to add or rotate keys; self-hosted and custom providers retain their Traffic Policy configuration.

### Source excerpt

Provider keys can now be attached directly to AI Gateway keys, removing the need for Traffic Policy config or a custom auth layer.

## What Is an AI Gateway, and Do You Need One?

DevFeed: [What Is an AI Gateway, and Do You Need One?](<https://devfeed.tech/articles/what-is-an-ai-gateway-and-do-you-need-one-83176.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/ai-gateways-2026>)

Author: Sam Richard

Published: 2026-04-09T00:00:00Z

Content type: article

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [gateway](<https://devfeed.tech/topics/gateway.md>), [Multi Agent Systems](<https://devfeed.tech/topics/multi-agent-systems.md>), [API Governance](<https://devfeed.tech/topics/api-governance.md>), [rate-limiting](<https://devfeed.tech/topics/rate-limiting.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>)

Tags: [access-control](<https://devfeed.tech/tags/access-control.md>), [agent-workflows](<https://devfeed.tech/tags/agent-workflows.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [ai-gateway](<https://devfeed.tech/tags/ai-gateway.md>), [ai-infrastructure](<https://devfeed.tech/tags/ai-infrastructure.md>), [api](<https://devfeed.tech/tags/api.md>), [api-gateway](<https://devfeed.tech/tags/api-gateway.md>), [api-keys](<https://devfeed.tech/tags/api-keys.md>), [costs](<https://devfeed.tech/tags/costs.md>)

### AI overview

The article argues that agentic AI has changed AI gateways from optional middleware for routing model calls into infrastructure for managing multi-step sessions. It describes using gateways to route among models, track agent behavior and costs, and enforce controls such as access policies, rate limits, and audit logging at MCP tool boundaries. It recommends gateways for teams running action-taking AI in production, while saying straightforward single-model applications may be able to skip them.

### Source excerpt

What is an AI gateway, and do you need one in 2026? How AI gateways route, secure, and observe LLM and agent traffic -- and when one is essential.

## Free plan users can now purchase AI Gateway credits

DevFeed: [Free plan users can now purchase AI Gateway credits](<https://devfeed.tech/articles/free-plan-users-can-now-purchase-ai-gateway-credits-83173.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/ai-gateway-free-plan>)

Author: Nijiko Yonskai

Published: 2026-04-08T00:00:00Z

Content type: release

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [litellm](<https://devfeed.tech/topics/litellm.md>), [transfer stations](<https://devfeed.tech/topics/transfer-stations.md>)

Tags: [ai-gateway](<https://devfeed.tech/tags/ai-gateway.md>), [api](<https://devfeed.tech/tags/api.md>), [api-keys](<https://devfeed.tech/tags/api-keys.md>), [model-pricing](<https://devfeed.tech/tags/model-pricing.md>), [model-routing](<https://devfeed.tech/tags/model-routing.md>), [ngrok](<https://devfeed.tech/tags/ngrok.md>), [pricing](<https://devfeed.tech/tags/pricing.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>)

### AI overview

ngrok now lets free plan users purchase credits and create AI Gateway API keys from the dashboard, so they can use ngrok-managed keys without first setting up provider accounts or upgrading. Provider costs are passed through, with a processing fee of $0.05 per million tokens for requests using managed keys; bring your own key remains free on all plans.

### Source excerpt

Purchase credits, create AI Gateway API keys, and route to multiple providers--even if you don't use the rest of ngrok... yet.

## Quantization from the ground up

DevFeed: [Quantization from the ground up](<https://devfeed.tech/articles/quantization-from-the-ground-up-83315.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/quantization>)

Author: Sam Rose

Published: 2026-03-25T00:00:00Z

Content type: tutorial

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Post-training optimization](<https://devfeed.tech/topics/post-training-optimization.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Ground truth / benchmark quality](<https://devfeed.tech/topics/ground-truth-benchmark-quality.md>), [NVFP4](<https://devfeed.tech/topics/nvfp4.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [awq](<https://devfeed.tech/tags/awq.md>), [benchmarking](<https://devfeed.tech/tags/benchmarking.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [llms](<https://devfeed.tech/tags/llms.md>), [quantization](<https://devfeed.tech/tags/quantization.md>)

### AI overview

This tutorial explains how quantization reduces the memory footprint of large language models by representing their parameters with fewer bits. It compares symmetric and asymmetric quantization, describes blockwise handling of outliers, and reviews ways to evaluate quality loss, including perplexity, KL divergence, benchmarks, and practical testing. In the reported tests, 8-bit and 4-bit versions retained useful performance, while 2-bit quantization substantially degraded the Qwen3.5 9B model.

### Source excerpt

A complete guide to quantization: how LLMs store parameters as floats, why shrinking them to 4 or 8 bits barely hurts accuracy, and how to measure the tradeoff.

## Pin your traffic to specific regions

DevFeed: [Pin your traffic to specific regions](<https://devfeed.tech/articles/pin-your-traffic-to-specific-regions-83164.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/2026-03-18-region-pinning>)

Author: Ben Kelley

Published: 2026-03-18T00:00:00Z

Content type: release

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Internet Traffic](<https://devfeed.tech/topics/internet-traffic.md>), [Network Segmentation](<https://devfeed.tech/topics/network-segmentation.md>), [Bot Management](<https://devfeed.tech/topics/bot-management.md>)

Tags: [ccpa](<https://devfeed.tech/tags/ccpa.md>), [costs](<https://devfeed.tech/tags/costs.md>), [data-residency](<https://devfeed.tech/tags/data-residency.md>), [gdpr](<https://devfeed.tech/tags/gdpr.md>), [latency](<https://devfeed.tech/tags/latency.md>), [networking](<https://devfeed.tech/tags/networking.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>), [regions](<https://devfeed.tech/tags/regions.md>), [security](<https://devfeed.tech/tags/security.md>), [traffic](<https://devfeed.tech/tags/traffic.md>)

### AI overview

ngrok has launched Region & IP Resolution, a self-serve feature that lets customers route a domain through selected geographic points of presence without changing DNS. It supports compliance and geographic policies while using latency-aware routing among chosen locations. The feature is available on the Pay-as-you-Go plan and costs $0.02 per active endpoint hour; ngrok says it plans to add regional aliases and dedicated IP addresses.

### Source excerpt

Your traffic, your rules. Route ngrok domains through specific geographic PoPs for compliance, security, or just because you want to.

## Secure internal Kubernetes connections with pod identity variables

DevFeed: [Secure internal Kubernetes connections with pod identity variables](<https://devfeed.tech/articles/secure-internal-kubernetes-connections-with-pod-identity-variables-83265.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/kubernetes-pod-identity-traffic-policy>)

Author: Brian Melton-Grace

Published: 2026-03-13T00:00:00Z

Content type: tutorial

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Access Control](<https://devfeed.tech/topics/access-control.md>), [Security, Identity, & Compliance](<https://devfeed.tech/topics/security-identity-compliance.md>), [Operator Lifecycle Manager](<https://devfeed.tech/topics/olm.md>)

Tags: [access-control](<https://devfeed.tech/tags/access-control.md>), [infrastructure](<https://devfeed.tech/tags/infrastructure.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [locking](<https://devfeed.tech/tags/locking.md>), [policy](<https://devfeed.tech/tags/policy.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>), [security](<https://devfeed.tech/tags/security.md>)

### AI overview

The ngrok Kubernetes Operator can expose pod identity metadata to Traffic Policy on Kubernetes-bound endpoints. Policies can use that identity to restrict service connections by workload, such as allowing only pods in a specified namespace. The article describes fail-closed and fail-open handling when identity cannot be resolved, and notes limitations for host-network pods and some CNI configurations.

### Source excerpt

Use Kubernetes pod identity variables in Traffic Policy to make access control decisions based on which workload is connecting.

## AI Gateway: use Anthropic and OpenAI with one API key

DevFeed: [AI Gateway: use Anthropic and OpenAI with one API key](<https://devfeed.tech/articles/ai-gateway-use-anthropic-and-openai-with-one-api-key-83172.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/ai-gateway-api-keys-credits>)

Author: Nijiko Yonskai

Published: 2026-03-02T00:00:00Z

Content type: release

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [gateway](<https://devfeed.tech/topics/gateway.md>), [API keys](<https://devfeed.tech/topics/api-keys.md>), [Claude](<https://devfeed.tech/topics/claude.md>), [OpenAI](<https://devfeed.tech/topics/openai.md>)

Tags: [ai-gateway](<https://devfeed.tech/tags/ai-gateway.md>), [api-keys](<https://devfeed.tech/tags/api-keys.md>), [cost](<https://devfeed.tech/tags/cost.md>), [failover](<https://devfeed.tech/tags/failover.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>)

### AI overview

ngrok has launched AI Gateway API Keys and prepaid credits, letting users send requests to OpenAI and Anthropic without creating provider accounts or managing provider keys themselves. The service validates the gateway key and supplies provider credentials, while users can combine managed keys with their own keys for other providers. Credits are prepaid, expire after a year, and cover provider costs plus a processing fee on requests using ngrok-managed keys.

### Source excerpt

Create an AI Gateway API Key, add credits, and start making requests to OpenAI and Anthropic--ngrok handles the rest.

## Lessons from my self-improving coding agent

DevFeed: [Lessons from my self-improving coding agent](<https://devfeed.tech/articles/lessons-from-my-self-improving-coding-agent-83198.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/bmo-self-improving-coding-agent>)

Author: Joel Hans

Published: 2026-02-25T00:00:00Z

Content type: article

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [agent orchestration](<https://devfeed.tech/topics/agent-orchestration.md>)

Tags: [agent-harness](<https://devfeed.tech/tags/agent-harness.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-coding](<https://devfeed.tech/tags/agentic-coding.md>), [agentic-loop](<https://devfeed.tech/tags/agentic-loop.md>), [ai](<https://devfeed.tech/tags/ai.md>), [coding-agent](<https://devfeed.tech/tags/coding-agent.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>)

### AI overview

The author describes building and using bmo, a coding agent designed to improve its tools and workflows through active learning, session reflection, and telemetry. In nearly 100 sessions, specialized tools helped make common tasks more reliable, but the agent often deferred improvements and could not independently sustain self-diagnosis. The author concludes that its gains come from prompts, tools, and scaffolding rather than changes to the underlying model, with human guidance remaining central.

### Source excerpt

Two weeks with a self-improving coding agent taught me more about LLMs than a year of prompting ever did.

## ngrok-operator v0.22.0: multiple installs and clean uninstalls

DevFeed: [ngrok-operator v0.22.0: multiple installs and clean uninstalls](<https://devfeed.tech/articles/ngrok-operator-v0-22-0-multiple-installs-and-clean-uninstalls-83297.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/ngrok-operator-multiple-installs-and-uninstall>)

Author: Alex Bezek

Published: 2026-02-25T00:00:00Z

Content type: release

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [Operator Lifecycle Manager](<https://devfeed.tech/topics/olm.md>)

Tags: [cluster](<https://devfeed.tech/tags/cluster.md>), [crd](<https://devfeed.tech/tags/crd.md>), [helm](<https://devfeed.tech/tags/helm.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [ngrok](<https://devfeed.tech/tags/ngrok.md>), [release](<https://devfeed.tech/tags/release.md>)

### AI overview

The ngrok-operator Helm chart v0.22.0 supports multiple operator installations in one Kubernetes cluster, separates CRDs into a standalone chart, and adds configurable uninstall behavior. Its Keep policy preserves ngrok API resources, while Delete removes them during cleanup.

### Source excerpt

This release unlocks multiple ngrok-operator installations per cluster, splits CRDs into their own chart, and adds a comprehensive uninstall system with configurable drain policies.

## Windows binaries are now signed with an EV certificate

DevFeed: [Windows binaries are now signed with an EV certificate](<https://devfeed.tech/articles/windows-binaries-are-now-signed-with-an-ev-certificate-83359.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/windows-ev-certificate>)

Author: Arianna Willett

Published: 2026-02-24T00:00:00Z

Content type: release

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [enterprise certificate authority certificates](<https://devfeed.tech/topics/enterprise-certificate-authority-certificates.md>), [certificates](<https://devfeed.tech/topics/certificates.md>), [Certificate Transparency](<https://devfeed.tech/topics/certificate-transparency.md>)

Tags: [certificate](<https://devfeed.tech/tags/certificate.md>), [digicert](<https://devfeed.tech/tags/digicert.md>), [digital-signatures](<https://devfeed.tech/tags/digital-signatures.md>), [fips](<https://devfeed.tech/tags/fips.md>), [microsoft](<https://devfeed.tech/tags/microsoft.md>), [microsoft-defender](<https://devfeed.tech/tags/microsoft-defender.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>), [security](<https://devfeed.tech/tags/security.md>)

### AI overview

ngrok has replaced the standard code signing certificate on its Windows executables with a DigiCert Extended Validation certificate. EV certificates establish immediate reputation with Microsoft Defender SmartScreen after stricter publisher identity checks and secure storage of the signing key. The change is intended to reduce warnings for fresh ngrok releases; the company notes that Defender behavior can still be unpredictable.

### Source excerpt

We've met stricter identity verification and hardware requirements for signing keys to stop Defender warnings without changing anything for you.

[Next page](<https://devfeed.tech/sources/ngrok-news.md?cursor=WyIyMDI2LTAyLTI0VDAwOjAwOjAwKzAwOjAwIiwgIjdhMDg1MjhkLWM2MWMtNDFkZS1hZjNlLWVlZmI2NjRkZTI5OCJd>)