Pricing

Billing Methods

Understand the different billing methods used across Skytells models and services.

Billing Methods

Each model and service on Skytells is priced differently depending on the resources it consumes. Below are the billing methods you may encounter.

Billing MethodDescription
Per SecondTime-based billing for processing services. Common for GPU-accelerated workloads.
Per UnitFixed price per API call or operation.
Per UsageBilling based on resource consumption (e.g., storage, bandwidth).
Per TokenCommon for language models. Billed separately by input tokens and output tokens.
Per PredictionBilled per prediction request. Common for image, video, and multi-modal models.
Per MegapixelBilled based on the resolution of generated images.

Per Second

Time-Based Billing

Models running on GPU or CPU hardware are billed by the second. The cost depends on the hardware tier selected. You only pay for the time your request takes to process.

This is the most common billing method for models that run on dedicated hardware. For example, a model running on an Nvidia H100 GPU is billed at a per-second rate for the duration of your request.

See Hardware Pricing for a full breakdown of per-second rates.


Per Token

Token-Based Billing

Language models are typically billed by the number of tokens processed — both input (prompt) and output (completion) tokens. Pricing varies by model.

Tokens are the basic units of text that language models process. A token is roughly 4 characters or about ¾ of a word in English. Input tokens (your prompt) and output tokens (the model's response) are priced separately.

You can find per-token pricing for each model on its page in Dashboard → Console.


Per Prediction

Prediction-Based Billing

Some models charge a fixed rate per prediction, regardless of processing time. This is common for image generation, video, and multi-modal models.

See Predictions for details on how predictions work and how they are billed.


Per Megapixel

Resolution-Based Billing

Image generation models may be billed based on the output resolution in megapixels. Higher resolution outputs cost more.


Per Unit & Per Usage

Some services use flat per-unit pricing (e.g., a fixed cost per API call) or usage-based pricing tied to resource consumption such as storage or bandwidth.


Where to Find Pricing

Each model displays its pricing and billing method on its page. You can browse all models and their costs from:


Need Help?

Contact Support

Submit a support ticket for billing or pricing questions.

How is this guide?

On this page