Skip to main content
The analyst is powered by a model: the AI that reads your question, looks things up, and writes the answer. The picker sits in the bottom-left of the chat composer, next to the reasoning-effort picker. Open it in the studio to see the current lineup; it changes as providers ship new models. Most questions are fine on the house model. Switch when you want a different model for speed, depth, or writing style.

​
Vela and frontier models

Hobby and Pro

Vela

Investi’s house model. Every plan includes a monthly allowance (higher on Pro). On Hobby it is the only model available. See Plans and billing for what happens at the cap.

Pro

Frontier models

Claude, GPT, Gemini, Grok, Muse Spark, Kimi, GLM, Qwen, DeepSeek, and others, on the Pro plan. On Hobby they appear with a “Pro” badge; selecting one takes you to Settings → Plans & Billing.

Every model in the picker runs the analyst and supports thinking.

​
Managing the picker list

Select Add Models in the picker footer (or go to Settings → Models) to choose which models appear in the picker. Each model has an enable/disable switch; at least one must stay enabled.

​
Reasoning effort

Reasoning effort is how hard the model thinks before it answers. Higher effort tends to produce more careful analysis at the cost of speed. On Hobby, Medium and above appear with a “Pro” badge; selecting one takes you to Settings → Plans & Billing.
For quick lookups, a low effort setting is faster and usually enough. Save higher effort for thesis work, comparisons, and anything you’ll act on.