Updates from ngrok
Developer infrastructure for routing and securing traffic to apps, APIs, and AI models.
- Gemini 4 Carbon is in AI Gateway, with a dial for how much it ponders (Product launch · November 11, 2026)
ngrok added support for Google's Gemini 4 Carbon reasoning model to its AI Gateway, allowing developers to configure thinking levels and manage costs via ngrok credits or their own Google API keys. - Run Liquid AI's small models and 160 more from LLM API in AI Gateway (Product launch · November 9, 2026)
The AI Gateway now supports Liquid AI's small models and over 160 models from LLM API, including Amazon Nova and Llama 3.3, which can be billed to ngrok credits or provider-specific keys. - Give your fallback models one name with virtual models (Product launch · October 29, 2026)
ngrok introduced virtual models for its AI Gateway, allowing users to group up to ten models under a single name to enable automatic failover without redeploying client code. - Ask @ngrok about your endpoints and traffic in Slack (Product launch · October 28, 2026)
The new AI Chat Agent for Slack allows teams to query their ngrok endpoints, traffic patterns, and account limits using natural language directly within Slack channels.