You.bot

You.bot is a pay-per-call API gateway that exposes 87 video, image, music and LLM models through one endpoint and one credit balance, priced below the direct provider rates it lists, with refunds on failed tasks.

You.bot is a multi-model API gateway that puts video, image, music and text models from several providers behind a single REST endpoint. Instead of keeping separate accounts, SDKs and billing relationships for each vendor, a team funds one credit balance and calls models from Anthropic, OpenAI, Google, xAI, DeepSeek and others through the same request format. The catalog lists 87 models, including Claude, GPT, Gemini, Nano Banana, Grok, Veo, Kling, Seedance and Suno. The target reader is a developer or product team building features that rely on generative AI, especially those that mix modalities. A typical app might need an LLM for copy, an image model for visuals and a video or music model for media, and the gateway lets all of them share one parameter schema. Changing models is described as a one-line change to the model id, which makes it practical to compare outputs or swap providers without rewriting backend code. In practice, a request is sent to the generate endpoint with a bearer API key, a model id and an input object. The call is asynchronous and returns a task id. Results arrive through a webhook, or the client can poll a task endpoint for state and result URLs. The site says fallback routing handles provider hiccups, and each model has its own page where prompts and parameters can be tried before any integration work begins. Pricing is usage-based and prepaid. Credits do not expire, there are no seats or minimums, and failed generations are refunded automatically. The listed rates sit below the public direct prices shown on the page, with a larger discount available through a $1,250 top-up that adds 10% bonus credits. Compared with calling each provider directly, the trade-off is an extra intermediary in the request path. Compared with other aggregators, the pitch rests on price and breadth across video, image, music and text rather than on tooling around prompts or evaluation.