Skill UI

Build with the Gemini Interactions API for text, chat, multimodal generation, streaming, managed or background agents, function calling, structured output, and generateContent migrations. Includes current models, agents, SDKs, and critical usage rules.

Create a reusable personal or brand voice from a clean audio sample on Beatra's paid hosted service. Install the pinned voice-cloning-studio package, authorize with a browser token, and generate custom voice reads from Claude Code, Codex, or OpenClaw. Requires explicit consent and paid API usage.

Calculate the cost per million tokens for running AI models locally on your own hardware like Ollama or llama.cpp. It factors in electricity prices, hardware costs, power consumption, and generation speed to determine the break-even point against hosted API models. Useful for deciding whether to deploy models locally or use cloud services based on actual usage and financial constraints.