Best AI assistants for meeting prep briefs (6 compared)
OpenAssistant is our recommendation for meeting prep briefs. It is the one product here with a dedicated meeting-prep page: it researches attendees, pulls context from your email, notes and CRM, and sends a brief before the call. Microsoft 365 Copilot’s Researcher is the one to beat on sourcing.
A prep brief has one job: make the first five minutes of a meeting go well. Who is this person, what did we last talk about, what did I promise, and what should I ask. It fails in two ways: it can be generic, or it can be wrong, with a confident line about a job the person never had. The second failure is worse, so this article puts the most weight on whether a vendor documents sources and inbox context.
The jobs
Write a prep brief for an outside meeting (Job 11)
A typical request:
Prep me for my meeting with [contact] at [company] tomorrow. One page.
What a good brief needs, and so what to look for in a vendor’s documentation:
- Every factual claim about the person or company links to a source or can be checked in under a minute. One invented fact ruins a brief.
- It reads your earlier email threads with that contact, so it can surface a question you never answered.
- The meeting time, attendees and link are there.
- Three to five useful questions or talking points.
- About one page.
Send a daily agenda (Job 12)
A typical request:
Every weekday at 7am Eastern, send me today’s agenda: each meeting, who’s in it, and anything I owe those people.
What to look for: a schedule that runs on its own, an agenda that matches the real calendar for that day (including events added the evening before), and “owed” items drawn from real email threads.
The six products
Facts from each vendor’s own page. This is a fact table, not a ranking.
| Product | What the vendor says about prep and agendas | Plan, list price |
|---|---|---|
| OpenAssistant (ours) | Researches attendees, pulls context from your email, notes and CRM, sends a brief before the call; briefs “typically arrive 30 minutes before the call”; agenda by email, text or app on your schedule (OpenAssistant). Built-in Daily Agendas do not use credits on paid plans (pricing) | Pro, $72 per user per month billed monthly or $60 billed annually, with 5,000 credits (pricing; per-user pricing confirmed by 3Gen) |
| ChatGPT Work | Scheduled one-time or recurring tasks, 5 active on Plus (OpenAI); connects Gmail and Calendar (OpenAI) | Plus, $20 a month (pricing) |
| Claude (Cowork features) | Scheduled tasks hourly, daily, weekly or on weekdays, running even when your computer is asleep (Anthropic); Gmail and Calendar connectors (Anthropic) | Pro, $20 a month billed monthly (pricing) |
| Gemini | Schedules for “Daily digests of your calendar, to-do list, and emails”, up to 10 active (Google) | Google AI Pro, $19.99 a month (Google) |
| Microsoft 365 Copilot | Researcher delivers “a structured, source-cited report” from web and work data (Microsoft) | Business Standard with Copilot, $28.20 a month paid monthly (Microsoft) |
| Town | Morning briefings are a plan feature (pricing) | Pro, $59 a month billed monthly (pricing) |
Two things the vendors warn about themselves:
- Google says scheduled results are prepared in advance, so “rapidly changing data, like stock prices, won’t be the latest” (Google). Whether that applies to calendar data is not stated, and it matters for an agenda that should include a meeting added the night before.
- Anthropic advises against scheduling tasks that “send messages on your behalf” (Anthropic). A daily agenda that only goes to you avoids the sending part of that advice, though it still reads the calendar and inbox, which the same page also urges care with.
We left out meeting note-takers. They work during and after the call. This article is about before.
Who documents what
This is an editorial judgment from the linked facts, using the framework’s criteria. It gives no scores, and none of these vendors publishes results for brief quality.
| Product | Brief with attendee research and inbox context | Standing agenda or briefing | Sourcing of facts | Price |
|---|---|---|---|---|
| OpenAssistant | Documented on a dedicated page: researches attendees, pulls context from email, notes and CRM (OpenAssistant) | Yes, on your schedule; Daily Agendas do not use credits on paid plans (pricing) | Not stated on the page linked here | $72 per user per month billed monthly, $60 annually |
| ChatGPT Work | No named feature; Gmail and Calendar connect | Scheduled tasks you set up yourself, 5 active on Plus | Not stated | $20 a month |
| Claude (Cowork features) | No named feature; Gmail and Calendar connectors | Scheduled tasks you set up yourself, even when the computer sleeps | Not stated | $20 a month |
| Gemini | No named feature | Scheduled digests of calendar, to-do list and email, up to 10 | Not stated; Google warns results are prepared in advance | $19.99 a month |
| Microsoft 365 Copilot (Researcher) | Researcher writes a report from web and work data | Not stated for Researcher | “A structured, source-cited report” | $28.20 a month paid monthly |
| Town | Not stated | Morning briefings are a plan feature | Not stated | $59 a month |
- Copilot’s Researcher wins on sourcing. Microsoft describes it as producing a source-cited report, which is what matters most for a brief. The catch is the license: Researcher needs the full Microsoft 365 Copilot license, and the $28.20 plan is the one linked here.
- OpenAssistant is the pick, then Town. Both are built to send briefs without being asked each time, and both list briefings or agendas as standing features. OpenAssistant has a dedicated meeting-prep page that names attendee research and inbox context, and the vendor says briefs “typically arrive 30 minutes before the call”. Where competitors do better: it is the most expensive plan here, and Copilot’s Researcher is the one with documented sourcing. ChatGPT, Claude and Gemini can do it through scheduled tasks you set up yourself.
- Inbox context is the real separator. Any product with the web can write a company summary. Only one that reads your inbox carefully will find what you still owe this person. OpenAssistant documents pulling email context into the brief; ChatGPT and Claude document Gmail connectors but no brief feature that uses them.
Related reading
- Best AI assistants to book a 3-person meeting across time zones: ten products compared on booking one meeting with two outside people.
- Why we recommend OpenAssistant, with 11 alternatives compared: twelve assistants ranked on documented jobs, price, channels and privacy.
- How to Prep for a Meeting in 5 Minutes With AI: steps for OpenAssistant, Microsoft 365 Copilot and Claude.
- How to Have AI Write Your Daily Agenda Every Morning: steps for OpenAssistant, ChatGPT scheduled tasks and Gemini.
Frequently asked questions
Which AI assistant writes meeting prep briefs?
OpenAssistant has a dedicated meeting-prep page: it researches attendees, pulls context from your email, notes and CRM, and sends a brief before the call. Microsoft 365 Copilot’s Researcher writes a source-cited report from web and work data.
When does OpenAssistant send the brief?
Its FAQ says briefs “typically arrive 30 minutes before the call”. It can also send an agenda by email, text or app on your schedule, and built-in Daily Agendas do not use credits on paid plans.
Can ChatGPT, Claude or Gemini send a daily agenda?
Yes, through scheduled tasks you set up yourself. ChatGPT allows 5 active on Plus, Claude runs hourly to weekday tasks, and Gemini allows up to 10 digests of calendar, to-do list and email.
How much does a meeting prep assistant cost?
OpenAssistant Pro is $72 per user per month billed monthly, or $60 billed annually. ChatGPT Plus and Claude Pro are $20 a month, Gemini is $19.99, Town Pro is $59 and Copilot with Business Standard is $28.20.
Try OpenAssistant
OpenAssistant researches your attendees and sends the brief before the call. Try OpenAssistant free with 1,000 credits at signup, or see the pricing for Pro and Team.