Models provided by qRaptor

You don’t need a provider account to start. Every subscription comes with a set of models qRaptor operates on your behalf — they appear in the same Configurations list as your own, marked with a System badge, and they’re selectable anywhere a model is.

Filter the list to Source → Provided by qRaptor to see just these.

The Configurations list filtered to platform-provided models, with the explanatory banner above the cards

What’s included

Which models you get depends on your plan — higher tiers include more capable ones. They span the same categories as your own configurations:

Models — language models for agent reasoning, chat and generation
Embeddings — for RAG, semantic search and memory
Voice — speech-to-text and text-to-speech
Image & Video — image and video generation

ℹ️

Image & Video is platform-only. These models come with your plan rather than being something you connect yourself, so that category contains no configurations of your own.

How they’re paid for

Platform models are billed in credits from your subscription. Nothing is charged to a provider account, because there isn’t one — qRaptor holds the relationship with the provider.

That is the main practical difference from a configuration of your own. Bring your own key and the provider bills you directly; use a platform model and it draws down credits. Both work identically from an agent’s point of view.

What you can’t do with them

They’re read-only. There’s no edit, no disable, and no way to see their credentials — the card carries a note saying as much. What you can do is choose where they’re used:

Pick one from any model selector when building an agent.
Set one as the default for a platform feature in LLM Settings.

Using your own account instead

If you’d rather bill your provider directly — or you need a model, region or fine-tune the platform doesn’t offer — create a configuration of your own and point the relevant use cases at it. That’s covered in LLM Settings under BYOK.

💡

It isn’t all-or-nothing. Many projects use platform models for most things and their own account for one or two — heavy code generation, say, or a model with a data-residency requirement.

Common issues

A model you expected isn’t listed — the set depends on your plan tier. Check your subscription, or use your own configuration instead.
You want to change a platform model’s settings — you can’t, but you can create your own configuration for the same provider and select that instead.
Something appeared that isn’t a model — platform infrastructure is filtered out of this list. If you see something you can’t use anywhere, tell support.

LLM Policies overview — putting guardrails around what agents send to a model.