📄️ Overview
Kertex AI is Kloudbean's fully managed generative-AI service — a Gemini-compatible API gateway with your own endpoint, API tokens, hard spend caps, usage metering, and OpenAI compatibility. Learn what it is and how it works.
📄️ Supported Models
The Kertex AI model tiers — Fast, Advanced, Light and Embeddings. Learn which Kloudbean-managed Gemini model each alias maps to, what each tier is best for, and how per-token pricing works.
📄️ API Tokens
Create, scope, and revoke Kertex AI API tokens. Learn how tokens authenticate your calls, what the kb_dev_/kb_stage_/kb_prod_ prefixes mean, per-token daily caps, scopes, and instant revocation.
📄️ Developer Guide
The Kertex AI developer guide: base URL and endpoints, authenticating with your token, the Google GenAI SDK (Python and Node), raw Gemini REST, streaming, token counting, embeddings, error handling, and monitoring usage and logs.
📄️ OpenAI Compatibility
Use Kertex AI from any OpenAI-compatible client or tool — OpenWebUI, LibreChat, Continue, LangChain and the OpenAI SDKs. Learn the /v1 base URL, how to authenticate, and how to connect OpenWebUI.
📄️ Configuration
How Kertex AI configuration works. Because Kertex is a transparent proxy, the generation settings your app sends through the SDK are passed to the model unchanged, while environment-level defaults are managed by Kloudbean.
📄️ Limits & Alerts
Control Kertex AI spend with daily and monthly caps, a per-minute rate limit, and a daily request cap — all enforced by the gateway in real time. Set threshold alerts by email and webhook, and send a utilization snapshot on demand.