Skip to main content
Pixio carries over 500 models. This page explains how the catalog is organized so you can find the right one quickly, instead of scrolling.
The complete, always-current list — every model, its inputs, its ranges, and its price — lives in the API reference and in the model picker inside Generate. This page is the map, not the inventory.

​
Organized by family

Models are grouped by the family that makes them. Pick the family first, then the specific model inside it. Families you’ll see:

​
Organized by capability

Every model declares what it converts. This is usually the fastest way to filter: Image and video dominate the catalog; text-to-image and image-to-video are the two largest groups by a wide margin.

​
What each model costs

Pricing is per model, and shown in the picker before you run anything:
  • Fixed — a flat credit cost per run, typical for images
  • Per second — video and audio models often scale with duration
  • By resolution or quality — higher settings cost more on many models
See Credits & Plans for how billing works.

​
What your plan can reach

Access is set per model:
  • Most of the catalog is available on any paid plan
  • Free accounts reach a large subset — enough to evaluate every category
  • Some models are free on specific plans, running without touching your balance
  • Maker Mode covers a wide set of models under unlimited generations
Models your plan can’t run are marked in the picker rather than hidden, so you can see what an upgrade would unlock.

​
Choosing a model

1

Start from the capability

Decide what you’re converting — text to image, image to video, video to video. That cuts 500 models to a few dozen.
2

Pick a tier

Most families ship a fast/cheap model and a flagship. Draft with the cheap one, finish with the flagship.
3

Check the inputs

Two models with the same capability can accept very different controls. Every input has a tooltip — read them before assuming a setting exists.
4

Test small

Short duration, lower resolution, one output. Confirm the model understands your prompt before spending real credits.
For video, generate the first frame as an image, then chain Create Video from it. You get far more control over composition, and you iterate at image prices. See chained actions.

​
New models

The catalog changes often. New models appear in the picker as soon as they’re enabled — nothing to install or update. Check the changelog for what landed recently, and Model Status if a specific model is misbehaving.

​
Custom models

Trained your own LoRA? It appears in the form for models that support custom weights. Trigger it with its trigger word in the prompt. Training requires Premium or above — see Model Training in the app.