cross-tool AI assistant
Notis
1,077
Entries in the public catalogue
current Website CMS inventory
Open research · Updated 2026-08-10
How much work can an AI assistant actually reach?
This benchmark separates a product's public promise from the tools and actions its public pages actually verify. It compares Notis with three adjacent products and publishes the source data for independent checking.
Two evidence scopes
Catalogue totals and sampled-page names are shown, not equated.
real-estate and mobile AI executive assistant
The Librarian
11
Names found on the cited page
distinct integrations visibly named on the cited homepage
personal AI advisor and reflection assistant
Kin
0
Names found on the cited page
no external app integrations visibly named on the cited homepage
AI efficiency advisory
Leania
Unknown
Comparable count unavailable
not publicly verifiable from the accessible source
These numbers are not an apples-to-apples market-share ranking. A zero means the cited page names no external app integration; it does not claim the product has none. “Unknown” means the accessible source was insufficient.
Verified action surfaces
Breadth is useful only when the assistant can act.
Notis
notis.ai- Read and write across connected work apps
- Run scheduled and event-triggered automations
- Use persistent memory and connected context
- Delegate multi-step work from conversational channels
WhatsAppTelegramSlackiMessagedesktopCLI
The Librarian
thelibrarian.io- Capture and qualify real-estate leads
- Match properties and retrieve market context
- Schedule appointments and automate follow-ups
- Sync contacts, calendars, email, notes, and selected CRMs
WhatsAppSMSiOSAndroid
Kin
mykin.ai- Persistent personal context
- Advisory-board perspectives
- Guided sessions and reflection
- Proactive insights and journaling
mobile appvoice
Leania
leania.aiThe accessible public source did not provide enough comparable product evidence. The dataset leaves the field blank instead of inferring.
Methodology
Designed to be checked, corrected, and reused.
- 01
Notis reports its current public catalogue inventory. Peers report only distinct names visible on the cited page; those sampled-page counts are labelled non-comparable, and unknown is reported instead of inferred.
- 02
A capability is marked verified only when the cited public product page describes the action or surface.
- 03
Referring-domain overlap comes from a bounded Semrush link-intersect sample collected through Monid. It is used to prioritize research and outreach, not as a complete backlink census.
Authority opportunity
The data is the linkable asset—not a generic comparison claim.
A bounded Monid/Semrush link-intersect found 50 referring domains across the four products. The published dataset exposes the methodology and caveats so journalists, founders, and researchers can cite or challenge the evidence.
Inspect the source data