Assistant

Updated May 14, 2026 · FinityAi

The assistant is in-app chat on our hosted service: streamed replies, your choice of model, and—when you allow it—tools across the integrations you connect. You can turn tool use off per chat for plain Q&A. Provider and API key setup lives under Integrations.

Turn it on and configure providers

Sign in, then open Integrations → Assistant to enable the assistant for your account and configure models and keys.
Choose whether you use FinityAi-hosted models or your own API keys where your workspace allows it, and add keys for the providers you rely on.
Browse the full catalog under Integrations — connect Notion, email, tasks, calendar, notes, messaging channels, and anything else your organization offers. Each integration has its own page with connection steps; deeper guides live in Documentation (for example Notion).
If your workspace supports optional web reading, turn it on under Integrations → Web reading and set crawl limits there. Availability depends on how your administrator configured the service.

Tools, integrations, and web reading

Two ideas work together: what you enable in Integrations for your account, and whether a given chat may use tools at all.

Integrations — Connect mail, Notion, tasks, calendar, documents, Obsidian, CRM, dev tools, and more when your workspace supports them. Until an app is connected (or a built-in feature like web reading is on), the assistant cannot act on it.
Connected apps & tools in chat — In Assistant settings, this switch controls whether the model may call tools for this conversation. Turn it off for chat-only replies even if integrations are connected.
Web reading — Loads public HTTPS pages you point to so the assistant can quote or summarize visible text. Logins, paywalls, and most authenticated flows usually do not work; use public listing or help pages when possible.

When messaging is set up for your workspace, channels such as Slack or WhatsApp can use the same integration tools as the web app—see Messaging & channels.

For Personal tasks, Personal notes, and Personal calendar (the hosted productivity apps), see Personal tasks, notes, and calendar.

Open chat

From the sidebar under Work, open Assistant, or go to /app/assistant. Use the sidebar to start a new thread or switch between threads. If the assistant is not configured, the app surfaces setup hints and links to Integrations.

Chat settings

In the assistant header, open Settings (gear icon).

Model — Choose Auto to let FinityAi select a suitable model for the task, or pick a specific provider and model from the options available to your workspace. If the live catalog fails to load, you may see common defaults until it recovers. For more providers, keys, or hosted-model options, use Integrations → Assistant.
Connected apps & tools — When on, the assistant may use apps you have connected. Turn it off for chat-only replies without tool calls.

Threads and where data lives

Conversation history appears in the assistant sidebar so you can return to recent work. Availability across devices depends on your workspace settings and account state; use the in-product controls for any export or deletion options that are available to you.

Usage and billing

For token or usage summaries tied to your account, open Token Usage under Account in the sidebar. Whether usage is billed through FinityAi-hosted inference or your own provider keys is determined on Integrations → Assistant.