Pricing Overview

Pricing Overview

Skytells uses a pay-as-you-go pricing model. You only pay for what you use — no upfront commitments, no monthly minimums.

Each model, service, and capability on the platform — whether accessed through the Playground, API, or SDKs — is priced independently based on the resources it consumes.

How Pricing Works

Most models are billed by the time they take to run. The price-per-second varies according to the hardware in use. When running or training a model, you only pay for the time it takes to process your request.

Some models are billed by input and output — for example, language models are commonly billed per token.

You'll find cost estimates for any model on the model's page in the Dashboard → Console.

Sample Model Pricing

Below are a few models across different billing units to illustrate how pricing varies:

Retrieve Pricing Programmatically

You can fetch current model pricing at any time via the REST API or any of the official SDKs.

REST API

curl https://api.skytells.ai/v1/models

Each object in the response includes a pricing field with the amount, currency, and unit.

TypeScript SDK

import { createClient } from 'skytells';

const client = createClient('your-api-key');
const models = await client.listModels();

models.forEach(model => {
  console.log(`${model.name}: $${model.pricing.amount} / ${model.pricing.unit}`);
});

Python SDK

from skytells import Skytells

client = Skytells("your-api-key")
models = client.list_models()

for model in models:
    print(f"{model.name}: ${model.pricing['amount']} / {model.pricing['unit']}")

SDKs are available for TypeScript, Python, Go, Ruby, and PHP.

Early Charges for Fraud Prevention

Sometimes, if your usage crosses certain limits for the first time, we'll charge you early. This helps us catch fraud before it becomes a problem. If that happens, we'll email you a receipt right away using the email address configured when you set up billing. The amount will show up as a credit on your next invoice.

Models That Call Other Models

When you run a model that calls other models, you're billed for the compute time of the root model and for any downstream models it calls. You can view the list of downstream models in the model's Pricing section.

Fast Booting Fine-Tunes

Sometimes, we're able to optimize how a trained model is run so it boots fast. This works by using a common, shared pool of hardware running a base model. In these cases, we only ever charge you for the time the model is active and processing your requests, regardless of whether it's public or private.

Fast booting fine-tunes are labeled as such in the model's version list. You can also see which versions support the creation of fast booting models when training.

Hardware & Run Time

Different models run on different hardware. You'll find the hardware specifications under the "Run time and cost" heading on each model's page in the Console. For a full breakdown of hardware tiers and their pricing, see Hardware Pricing.

Explore Pricing

Need Help?