Gemini 4 Carbon is in AI Gateway, with a dial for how much it ponders
Call 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.
34 entries
Call 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.
Call Anthropic's newest Fable model on your ngrok credits or your own Anthropic key, at $8 input and $40 output per 1M tokens.
New models from Moonshot AI, OpenAI, MiniMax, Tencent, and Google run on your ngrok access key and credits, and switching between them changes one string.
Two 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.
Lyria 3.5 writes full songs and Lyria 3 Clip Preview makes 30-second clips, at a flat price per song on your ngrok credits.
Turn text into speech with Gemini, xAI, MiniMax, and Qwen voices on your ngrok credits, or bring your own Groq and Mistral keys.
Save an ordered list of models as a virtual model, send its name as the model, and AI Gateway tries each model in turn until one answers.
Connect Slack to your ngrok account, then mention @ngrok in a channel to ask about your endpoints, traffic, usage, and limits and get the answer in a thread.
Four Gemini image models now draw and redraw pictures on your ngrok credits, from your code or the playground.
Attach a Firewall policy to your access keys to check what agents send to a model and stop risky tool calls before your agent runs them.
The dashboard Limits page now shows each limit's status, your current usage, and when it last rejected a request, starting with 11 account resources.
Traffic Inspector now lists TCP connections next to HTTP requests, including connections to TCP and TLS endpoints.
Turn text into vectors with OpenAI, Gemini, Qwen, Tencent, and Mistral embedding models, using your ngrok access key and credits.
Install an official ngrok package for Next.js, FastAPI, Axum, or Gin to give your app a public ngrok endpoint when it starts.
Turn on automatic top-up, and AI Gateway refills your balance from your saved card when it runs low.
Send Gemini Interactions API requests through AI Gateway with Google's Gen AI SDK for Python or TypeScript.
When your bank wants to confirm a credit purchase, AI Gateway gives you a link to approve it, and your credits arrive once you do.
AI Gateway now rejects requests larger than 32 MiB with a clear error, before they reach a provider.
When a model can't use a setting you sent, the Ngrok-AIG-Removed-Params response header names it.
Streaming responses now run for up to 15 minutes, so long answers and slow reasoning models finish.
Call OpenAI's GPT-6.1 Sol through AI Gateway with your ngrok access key and credits.
Call Anthropic's Claude Opus 5.5 and OpenAI's GPT-6 Sol and GPT-6 Luna through AI Gateway with your ngrok access key and credits.
Call five xAI Grok models, including grok-4.3 with a 1M token context window, through AI Gateway with your ngrok access key and credits.
Try the latest GLM additions with your ngrok credits, without creating separate provider accounts or managing their API keys.
Cursor users can now find and install ngrok's coding agent skills straight from the plugin marketplace, without cloning the repo.
Install eight new skills that teach Claude, Cursor, and other coding agents to secure endpoints, receive webhooks, test MCP servers, and troubleshoot ngrok.
Use Jev by Typesafe AI in AI Gateway to turn application data into typed decisions with calibrated probabilities, billed directly to your ngrok credits.
Copy a setup prompt from the dashboard so your coding agent can start ngrok for your endpoint and check the connection.
Attach images and text files, inspect reasoning and tool calls, and retry individual models while comparing responses in the playground.
Use your ngrok access key and credits to call supported DeepSeek, MiniMax, and QwenCloud models through AI Gateway.
Look up one request by ID to see each attempt the gateway made, and read which fallback models never ran.
Every AI Gateway response now carries its request ID, its attempt count, and the model that served it.
A policy with the ai-gateway action now fails validation with ERR_NGROK_3820, so send LLM traffic to gateway.ngrok.ai with an access key.
AI Gateway has its own dashboard at app.ngrok.ai, and your apps send every LLM request to gateway.ngrok.ai with one access key.