July 2026 Summaries

The rebuilt ngrok Terraform provider, now live as v0.8.1, facilitates the management of ngrok infrastructure through declarative, importable, and reviewable resources, which are integrated with version control workflows. This approach allows teams to manage domains, cloud endpoints, access policies, and credentials alongside their services, streamlining changes through PR reviews and eliminating manual credential sharing. The provider supports cloud endpoints paired with Traffic Policy, enabling control over authentication, rate limiting, and routing through code, thereby ensuring consistent policy enforcement across multiple endpoints. Credential management is enhanced with ngrok_vault and ngrok_secret, making credentials first-class Terraform resources and eliminating the need for out-of-band secrets. Existing infrastructure can be brought under management using terraform import, which allows teams to adopt Terraform without reprovisioning, thus easing the transition for those already utilizing ngrok in production environments.

Ngrok has implemented several pricing changes aimed at making its infrastructure services more cost-effective and aligned with actual usage, benefiting most users. Key changes include a significant reduction in Traffic Processing Unit (TPU) pricing, from $1 to $0.10 per 100,000 units, making it more affordable to implement robust protections like blocking bots and enforcing authentication. Custom domains will now only incur charges during active traffic, eliminating costs for idle domains, which is advantageous for those with variable or seasonal traffic patterns. The adjustment to Endpoint Pooling will result in billing for each active endpoint separately, though idle endpoints remain free, impacting a small fraction of heavy users. These changes reflect ngrok's commitment to providing infrastructure that supports developers' needs in production environments without excessive costs, with the company open to feedback and adjustments as necessary.

ngrok.ai has revamped its AI Gateway to offer a more streamlined experience for developers creating AI-powered applications, addressing previous complexities and focusing on centralized management and routing. The new platform, accessible via app.ngrok.ai, provides a dedicated dashboard, API, and gateway URL, allowing users to send requests to both hosted models like OpenAI and local models through a single endpoint. This setup enables seamless integration and easy switching between different models by changing the model name. Users can now configure access keys for specific environments or teammates, ensuring controlled and private model access while maintaining visibility over usage and spend. Furthermore, the new system supports both ngrok.ai credits and existing provider keys, offering flexibility in managing costs and provider relationships. The updated experience includes a small processing fee and requires users to transition from the old dashboard, with ngrok.ai offering assistance in moving any existing credit balances.