Notion AI

Free tier

AI writing and thinking assistant built directly into Notion's docs and databases.

Free tier available·All audiences·Powered by Model agnostic (multiple LLM providers)

Key strengths

AI agents (Notion Agent & Custom Agents) for autonomous task automationEnterprise Search across connected apps (Slack, Google Drive, GitHub)AI Meeting Notes with transcription and summariesModel-agnostic architecture (switch LLM providers without losing context)Zero data retention for Enterprise with LLM providersSOC 2 Type 2, ISO 27001, GDPR, CCPA, and HIPAA complianceGovernance tools and admin controls for AI usage

Free tier + paid plans · from $10 USD/mo

San Francisco, US

Founded 2016

Facts verified against notion.com on Oct 8, 2026

Recent changes

Oct 8, 2026Free tier: now available
Oct 8, 2026Plans updated: Free plan at $0 (trial AI only); Plus at $10/member/month (trial AI only); Business at $20/member/month (full Notion AI included); Enterprise at custom pricing. Custom Agents add-on: $10 per 1,000 monthly Notion credits.
Oct 8, 2026Models updated: Model agnostic (multiple LLM providers)

What's changed across AI tools →

Get alerts when Notion AI changes

Free. We'll email you when pricing, the free tier, API access or open-source status changes.

No ratings yet

Notion AI — Technical Documentation Notes

Pricing & Access

Plan	Price	AI Access
Free	$0	Trial AI only
Plus	$10/member/month	Trial AI only
Business	$20/member/month	Full Notion AI included
Enterprise	Custom pricing	Full Notion AI + advanced controls
Custom Agents add-on	$10 per 1,000 monthly Notion credits	Autonomous agent execution

AI Agent Architecture

Notion Agent: Built-in autonomous agent for task automation within the Notion workspace.
Custom Agents: Configurable agents for bespoke automation workflows, billed via Notion credits.

Integrations

Notion AI's Enterprise Search indexes and queries across: Slack, Google Drive, GitHub, Asana. Additional integration support may vary by plan.

Model Provider

Notion AI is model-agnostic — it interfaces with multiple LLM providers. The abstraction layer ensures workspace context and data are preserved when providers are switched.