> For the complete documentation index, see [llms.txt](https://docs.nusa.nissology.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.nusa.nissology.ai/renting/renting-a-model.md). # Renting a model On Nusa you rent a whole machine by the hour, for a model you choose before you pay. This page walks through choosing a model, choosing a machine, and paying. You don't need an account to look around. You can browse every model and machine, and see what a rental would cost, before you sign in. Nusa asks you to sign in when you choose to pay. To pay, you need Credit (see [Credit](file:///6919936/credit/credit.md)). ## Step by step {% stepper %} {% step %} ## Choose a model Open **Models**. You can search by a model's name, its brand, or what it's good for, and narrow the list to one use, such as chat, coding or reasoning. Each model shows what it's good for, its size, the longest context it supports (see [Context length](#context-length)), the GPU memory it needs, and a starting price per hour in USD. The starting price is the hourly price of the cheapest machine that can run the model. That machine may be rented by someone else right now, so the price you end up paying can be higher. {% endstep %} {% step %} ## Choose a machine Once you choose a model, you see the machines that can run it right now. If one meets Nusa's recommendation, it's marked Recommended and chosen for you, and you can see why. Otherwise, choose one yourself. You can sort the machines by lowest price, highest reliability score, or nearest. Nearest means where the machine is, not that it's faster. Each machine shows: * its GPU model and GPU memory * the country it's in * its reliability score, from 0 to 10: how well its host keeps it running for renters * how many hours it has been rented in all * its hourly price * the longest context it allows * whether it already has the model installed, or roughly how long installing it takes Your own machines aren't offered: an account can't rent a machine it hosts. {% endstep %} {% step %} ## Choose how long At checkout, enter how long you want the machine in days and hours. Rentals are charged in whole hours, and last at least one hour. Checkout starts at 24 hours, and you can change it. During Beta, a rental can be up to 7 days. A rental can't be extended. The total is the hours you entered multiplied by the machine's hourly price. It's exactly what you'll be charged. {% endstep %} {% step %} ## Check the rental Checkout shows everything about the rental before you pay: the model, the machine's GPU and the country it's in, how long and when it ends, the total, and how long that price is held (see [How long a price is held](#how-long-a-price-is-held)). It also shows the refund promise: **"Full refund if your rental ends unexpectedly."** (see [Refunds](broken://pages/dd006d856fc6112506772b72512b8f2fa0fc8285)). If the machine doesn't have the model installed yet, checkout shows roughly how long installing it will take and how much of your rental that uses. Installing counts as part of your rental. {% endstep %} {% step %} ## Pay Choose to pay. If you haven't signed in, Nusa asks you to now, then brings you back to checkout with your choices kept. Before anything is charged, you see the machine, the model, how long and the total once more. You're charged only when you confirm. The total is paid from your Credit (see [If your Credit isn't enough](#if-your-credit-isnt-enough)). {% endstep %} {% step %} ## Start using it Once you've paid, you go straight to your rental, with nothing to start. If the machine is installing the model, you see its progress. You can chat with the model in the **Playground**, or call it from your own code with the details in **Connect**: the base URL, your rental's key, the model's name and an example call. The key stops working when the rental ends. You can also switch to another model the machine can run during the rental. Each rental is a **job**. You find all your rentals in **Jobs**, where active ones come first. {% endstep %} {% endstepper %} ## Renting from the Marketplace You can also start from the machine instead of the model. The Marketplace lists machines, which you can narrow by price, GPU memory, machine type, continent, or a search. On a machine's own page, choose the model to run on it, then go on to checkout as above. ## Context length A model's context length is how much text it can take in at once: your prompt, the conversation so far and its answer, together. A machine with less GPU memory than the model needs can still run it, with a shorter context. It's marked as running the model with a shorter context, and shows how long. Checkout shows the longest context for your rental. ## How long a price is held The price shown when you confirm is held for five minutes. * **The price is held, not the machine.** Someone else can still rent the machine meanwhile. If they do, you aren't charged. * **If the host changes the price** within the five minutes, you pay the price you were shown. * **If the five minutes run out**, nothing is charged. You're shown the price again and confirm it. * **Once you've paid**, a later price change never affects your rental. ## If the machine is no longer available A machine can stop being available between choosing it and paying, for example because someone else rented it or it went offline. If that happens, Nusa tells you and offers other machines for the same model. The duration you entered is kept, and you confirm the new total before you're charged. If no machine can run the model right now, Nusa says so and helps you choose another model. ## If your Credit isn't enough Checkout shows how far your Credit is from the total before you choose to pay. Topping up isn't open yet. To get more Credit, [contact Nusa support](mailto:support@nissology.ai). --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.nusa.nissology.ai/renting/renting-a-model.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.