- Install ngrok
- Set up your coding agent
Install the ngrok agent and connect it with your authtoken.
On Linux or Windows, get the agent from the download page instead.The last command puts the service on that port online, whether it runs on your laptop or a production server.
Pick a job below to secure it, route it, or connect more services.
What do you want to do?
Run production traffic
Connect customer networks
Reach APIs and databases you don’t host.
Secure an API
Authenticate, rate limit, and route requests to your API.
Reach devices in the field
Connect to devices without opening inbound ports.
Receive webhooks on-prem
Verify webhooks, then route them to private services.
Expose Kubernetes services
Add ingress to any cluster with ngrok.
Open SSH and RDP sessions
Reach remote shells and desktops without a VPN.
Call AI models
Make your first request
Call OpenAI or Anthropic without a provider account.
Use a model you run
Route to Ollama or vLLM on hardware you control.
Bring your own keys
Bill requests to your provider accounts, not credits.
Build and test on your machine
Gather what you need
Every quickstart starts from an ngrok account. Most also ask for one of these:
The ngrok MCP server signs in with OAuth, so it needs only your account.
Choose how you run ngrok
Every product works with the same tools, so pick the one that fits where your service runs.CLI
Run
ngrok on any machine or device.SDKs
Embed ngrok in Go, JavaScript, Python, or Rust.
Kubernetes Operator
Use Ingress, Gateway API, or CRDs.
Cloud Endpoints and the API
Configure endpoints in the dashboard, API, or Terraform, with no agent.
Get help
FAQ
Answers to account, billing, and platform questions.
Error codes
Look up any
ERR_NGROK code and its fix.Pricing and limits
See what each plan includes and how usage is billed.
From the changelog
November 11, 2026Gemini 4 Carbon is in AI Gateway, with a dial for how much it ponders
View all updates · Subscribe via RSSCall Google’s newest reasoning model on your ngrok credits or your own Google key, and pick a thinking level for each request to control what an answer costs.
November 10, 2026Claude Fable 5.2 is in AI Gateway, and it costs less than Fable 5.1Call Anthropic’s newest Fable model on your ngrok credits or your own Anthropic key, at $8 input and $40 output per 1M tokens.
November 9, 2026Run Liquid AI’s small models and 160 more from LLM API in AI GatewayTwo new providers add Liquid AI’s small LFM3 models and more than 160 models from LLM API, including Amazon Nova, Llama, and ElevenLabs voices, on your ngrok credits or your own keys.