What model does Inoreader Intelligence use by default?

Inoreader Intelligence uses OpenAI's gpt-4o-mini by default. OpenAI is the default platform, and each platform has a default model of its own: mistral-small-2506 on Mistral and claude-haiku-4-5-20251001 on Claude (Anthropic).

The default model runs on Inoreader's own account with that vendor and uses your monthly Intelligence tokens. It handles article summaries and questions, rules that summarize, AI summaries in email digests, Intelligence reports and automated reports.

See the model in use

1. Click the gear in the tab bar and open Intelligence.
2. Open the Platforms tab. The selected card is your platform, and the line under its name is the model in use.

Selecting another card switches the platform and with it the model, and the change is saved at once. Which AI platforms does Inoreader support? describes the three choices.

Features with fixed models

Some features use the same model whatever platform you select. Suggested tags run on Google's gemini-2.5-flash-lite, podcast and video topics and article versions run on OpenAI's gpt-4.1-mini, and the AI overview on a mention's page runs on OpenAI's default model.

Choose a different model

To run another model, add your own API key. The card then shows the model you picked instead of the default, and requests run on your account with that vendor. Can I use my own API key? explains how. On a Team plan, the team's card uses the model the owner picked with the team's key, or the platform's default when there is none, as Can my Team have a default AI platform and model? describes.

Plans

Choosing a platform is part of Intelligence on Pro and higher plans, and choosing a different model needs your own API key with the Intelligence reports add-on or a Team Intelligence plan. The blog post Choose your AI platform for Intelligence features introduces the platform choice.