What did Hlido score joinly.ai?

joinly.ai scored 68/100 (FADING) in Hlido's independent, hands-on review.

Does any vendor pay Hlido for placement?

No. Hlido takes no money from the agents it rates — scoring weights stay private and the evidence behind every verdict is public.

joinly.ai

Name: joinly.ai review
Item: joinly.ai
Rating: 68
Author: Hlido Editor

Productivity · tested 2026-08-21 · re-test due 2026-11-21 · by the Hlido desk, not the vendor

In short: Meeting agents that actually participate — an AI that listens, writes in the chat, and speaks back during the call, exposed as an MCP interface — a genuinely different take on meeting AI that's still early.

4 PASS · 0 FAIL of 4 public-surface claims

Visit joinly.ai → scorecard.json

Quick answer

joinly.ai scores 68/100 (FADING) on Hlido’s independent, hands-on test (reviewed 2026-08-21). FADING (68, just below STEADY) for a genuinely differentiated, MCP-native approach to meeting AI (an agent that participates rather than just transcribes, and is itself an interface other agents plug into) with real evid

Hlido Editor · 2026-08-21

joinly.ai reframes meeting AI away from passive note-taking toward active participation: an agent that writes into the meeting chat, listens and speaks back interactively during the call, and does work in-meeting (web search, drafting presentations, updating a Miro board, follow-up emails, documents). The interesting architectural bet is that joinly is 'the meeting interface for your AI agents' — MCP is used inside the product and you can connect external MCP servers to customise the meeting agent — which fits the agent-to-agent thesis better than closed notetakers do. There's a real blog documenting the MCP backbone and a 'digital twin that joins low-value meetings for you' tutorial, signalling genuine building. It lands as FADING rather than STEADY because it reads as early and unproven: key capabilities like screen-sharing are marked 'Soon', there's no pricing/scale/reliability evidence on the surface, and real-time speak-and-listen-in-meeting is a hard reliability problem whose quality can't be judged from the page. Differentiated and worth watching for teams comfortable with early tools; verify latency, transcription accuracy and stability in your own meetings before relying on it.

Why FADING

FADING (68, just below STEADY) for a genuinely differentiated, MCP-native approach to meeting AI (an agent that participates rather than just transcribes, and is itself an interface other agents plug into) with real evidence of building (blog on the MCP backbone, digital-twin tutorial). Not lower because the concept and MCP-first architecture are real and distinctive. Not STEADY because it's early — a flagship feature is 'Soon', there's no pricing/scale/reliability evidence, and interactive real-time speak/listen quality is exactly what a homepage can't demonstrate.

Public-surface checklist

PASS Homepage loads (required)
PASS Primary value prop (required) — Meeting agents that actually participate — an AI that listens, writes in the cha
PASS Cta present
PASS Evidence or demo — 4 screenshot(s) captured

What we saw

4 screenshots captured by the Hlido engine during the reviewed run (run-15b724c7c5b47f55-joinly-ai). Our own captures — not vendor marketing material.

joinly.ai — run screenshot 1 (home.png) — home.png

joinly.ai — run screenshot 2 (page_.png) — page_.png

joinly.ai — run screenshot 3 (page_blog_mcp-the-backbone-of-joinly_.png) — page_blog_mcp-the-backbone-of-joinly_.png

joinly.ai — run screenshot 4 (page_blog_digital-twin-tutorial_.png) — page_blog_digital-twin-tutorial_.png

What it does well

Genuinely differentiated: an agent that participates in the meeting (writes in chat, listens, speaks back) rather than only transcribing afterward
MCP-native architecture: joinly uses MCP internally and lets you connect external MCP servers to customise the meeting agent — a real agent-to-agent surface
Concrete in-meeting actions: web search, drafting presentations, Miro-board updates, follow-up emails, document creation
Evidence of real building: blog on 'MCP: the backbone of joinly' and a 'digital twin that joins low-value meetings' tutorial

What it fails at

Early and unproven: a flagship capability (screen sharing) is explicitly marked 'Soon'
No pricing, scale, adoption or reliability evidence on the public surface
Real-time listen-and-speak in meetings is a hard latency/accuracy problem whose quality can't be judged from the page
Interactive meeting participation raises consent/recording considerations not addressed on the surface

Red flags

The differentiator — real-time interactive speak/listen participation — is the hardest thing to get right and is unverifiable from the surface; test latency, transcription accuracy and stability in real meetings first.
A headline capability (screen sharing) is marked 'Soon', so the current product is narrower than the pitch implies.

Best for

Teams wanting an AI that actively participates in meetings (searches, drafts, updates boards) rather than just taking notes
Developers who want a meeting agent they can extend by connecting their own MCP servers
Early adopters comfortable trialing a young, differentiated product

Not recommended for

Buyers who need a proven, reliable notetaker with pricing and scale evidence today
Environments with strict meeting-recording/consent policies (not addressed on the surface)
Anyone who needs the 'Soon' features (e.g. screen sharing) now

Compared to

MeetGeek | AI Note Taker and Meeting Assistant interactive-mcp-native-meeting-agent-vs-mature-hosted-notetaker
MeetGeek is a mature, hosted meeting-notetaker focused on recording, transcription and summaries; joinly.ai bets on active in-meeting participation and an MCP-native, agent-extensible architecture instead. Choose MeetGeek for proven notetaking today, joinly for a differentiated (but earlier) interactive meeting agent you can extend.

Agent relevance

API MCP

Agentic-Commerce Readiness 55/100 · INTEGRABLE

Independent readiness for agent delegation & transaction. How it’s scored · check live

Positioned as 'the meeting interface for your AI agents': MCP is used internally and external MCP servers can be connected to customise the meeting agent, so agents can act inside live meetings. Agent-to-agent-relevant by design, but real behaviour depends on live meeting infrastructure and can't be exercised from the public surface.

Agent-friendly score: 7/10

Evidence

Make meetings accessible to AI agents: brainstorm, search, draft presentations in-meeting — source (2026-08-21) verified
Writes with you (into meeting chat), speaks with you (interactive listen/speak), screen sharing marked 'Soon' — source (2026-08-21) verified
In-meeting capabilities: web search, presentations, Miro-board updates, follow-up mails, documents — source (2026-08-21) verified
MCP-native: joinly uses MCP internally and supports connecting external MCP servers — source (2026-08-21) verified
Digital-twin tutorial: create a twin that joins low-value meetings for you — source (2026-08-21) verified

scorecard.json · transparency passport · registry · methodology

More: compare agents · best of · developer tools · incident registry

Verdict by Hlido Editor, our automated editorial system · Method: public-surface-tier-1+editorial-narrative-v2 · Methodology version 2026.05 · Next review due 2026-11-21

How this page was produced. The scores, claim verdicts and evidence come from automated hands-on testing of the product’s public surface. The written analysis is drafted by an AI system, and pages publish without a person reviewing each one. Hlido publishes this record and answers for it — tell us if anything here is wrong and we will correct it.

Embed this trust badge

Live, always-current independent score — free to embed on your site or README. No vendor pays for placement.

Markdown

[![Hlido trust score](https://hlido.eu/badge/joinly-ai-joinly.svg)](https://hlido.eu/check/?agent=joinly-ai-joinly)

HTML

<a href="https://hlido.eu/check/?agent=joinly-ai-joinly"><img src="https://hlido.eu/badge/joinly-ai-joinly.svg" alt="Hlido trust score"></a>