Models

The Models view lists every model available to your organization and lets you decide which are enabled. Each row shows the model’s type and the sub-processor that actually receives your data when it runs, which is the column that matters most for a DPA review.

The four model types

• Completion models — the engine behind an assistant’s reasoning and text generation.
• Embedding models — convert documents into searchable vectors for knowledge search.
• Transcription models — convert speech to text.
• Image models — generate and edit images from text descriptions.

The sub-processor column

Each model’s row shows its sub-processor: the model vendor for a model accessed through an external API, or Intric’s own infrastructure provider for a model Intric hosts itself. Use this column when reviewing Data Processing Agreements or deciding which models suit a given security classification.

Curated picks

Beyond individual models, Intric offers four curated picks assistants can use instead of naming a specific model: Balanced, Fast, Economical, and Smartest. Each one tracks a good current model for that priority as the underlying catalogue evolves.

An assistant set to a pick is bound to the pick rather than to the model behind it on the day it was created, so it follows the catalogue as you change what’s enabled here. The assistants table shows each one’s pick together with the model it currently resolves to.

Ratings for custom models

A custom completion model can be bound to a benchmark Intric has synced, so its speed and intelligence ratings follow that benchmark instead of figures typed in by hand — and stay right as the benchmark is updated. Intric suggests a matching benchmark from the model’s name when you add it; accept the suggestion, pick a different benchmark, or enter the figures manually as before.

Token cost for custom models

Important

For standard vendor models, Intric already knows the pricing and applies it automatically to usage and cost reporting. For custom models — self-hosted or organization-specific deployments — Intric has no way of knowing the price, so set the token cost yourself on the model’s configuration, matching your actual negotiated rate rather than a public list price. Keep it current: an outdated or missing value leads to under- or over-reported figures wherever that model’s cost is shown, including Usage and Billing.

How to choose which models to enable

A few principles worth applying when deciding which models to enable and where:

• Most modern models are general-purpose and handle everyday tasks well — don’t over-complicate the choice.
• Use a fast model for the personal assistant and simple tasks; reserve models with stronger reasoning for complex analysis.
• Limit access to the most capable — and often most expensive — models to the classifications where they’re genuinely needed.