Meeting minutes
Somebody always volunteers to take notes in a meeting, and the notes are always incomplete because taking notes and participating are two jobs fighting for the same attention. This endpoint takes the notes so your team can just talk.
Run it online
Run this on our servers with your account. Free tools run in your browser; this one bills your KIT balance per the price above.
The actual job of meeting minutes
Good minutes are not a transcript with the small talk cut out; they are a structured record of what the meeting was for — what got decided, who owns what, and what happens before the group meets again. Most manually-taken notes drift toward one extreme or the other: too sparse to be useful later, or too long for anyone to reread. audio.meeting_notes is built around that structure from the start, not bolted on after the fact.
How the request works
POST your recording to /audio/meeting-notes and get a task_id back immediately; the meeting is processed off the request thread and delivered to your webhook, or via a signed link valid for 24 hours, once it's done. The output is organized around decisions, owners and next steps rather than a flat wall of text, so someone catching up later can scan it in under a minute instead of reading it top to bottom.
Why owners and dates are the hard part
Extracting a decision from a conversation is comparatively easy; correctly attributing who owns the follow-up and by when is the harder problem, because meetings are full of tentative commitments, half-agreements and things people say they'll 'look into.' We surface what the recording actually supports rather than inventing a due date that was never stated, and where ownership is ambiguous the notes reflect that ambiguity instead of quietly guessing.
What it assumes about your meetings
It works best on meetings that were actually discussions — status updates, planning sessions, client calls, retros — where decisions and commitments are spoken aloud, as opposed to meetings that are really just one person presenting slides for forty minutes. Multiple speakers, some crosstalk and a normal amount of tangent are all expected and handled; a completely silent screen-share is not going to produce useful minutes because there's nothing spoken to extract.
Where it plugs into your workflow
Teams commonly chain it straight into a task tracker or project management tool by feeding the structured output into whatever consumes action items downstream, or forward the notes to attendees automatically after every recurring call. Because it's billed per request plus per minute and only on success, it scales the same way whether it's one weekly stand-up or a full day of back-to-back client calls.
What you can do with it
Weekly stand-up records
An engineering team auto-generates minutes from its recorded stand-up and posts decisions to the team channel without anyone typing notes.
Client call documentation
An agency turns every client call recording into structured minutes that get attached to the project file automatically.
Board and committee meetings
A nonprofit produces a first-pass draft of meeting minutes from the recording, which the secretary then reviews and finalizes.
Cross-timezone handoffs
A distributed team shares structured minutes instead of a full recording so a colleague in another timezone can catch up in minutes, not an hour.
FAQ
How do I generate meeting minutes from a recording via API?
Send the recording to POST /audio/meeting-notes, keep the returned task_id, and receive structured minutes by webhook or a signed link valid for 24 hours.
Is the meeting transcription API free?
No, there is no free tier or trial. It's priced at $0.004 per request plus $0.0185 per minute, and a failed task is never billed.
What does the output actually contain?
Minutes organized around decisions made, owners assigned to follow-ups, and next steps, rather than a raw transcript.
Does it identify who said what?
It attributes decisions and action items to speakers when that's clear from the audio; for guaranteed speaker labels throughout, pair it with a diarized transcription task.
Can it handle meetings with many participants?
Yes, multi-speaker meetings with normal crosstalk are the expected case, though very large calls with many simultaneous voices are harder for any system to parse cleanly.
What if no clear decisions were made?
The notes reflect that honestly rather than inventing decisions or owners that the recording doesn't support.
How long can the meeting recording be?
It processes asynchronously specifically to handle long meetings without timing out; you're notified once minutes are ready regardless of length.
Is the recording stored afterward?
No, audio and generated minutes are deleted after the retention period and are never used to train models.
For developers — API access
Everything on this page is available programmatically. This section is for teams who want to wire it into their own systems; everyone else can just use the tool above.
API endpoint
Prefer to automate it? One authenticated POST creates the task; the result comes back by webhook or a signed link. The same capability also runs here on the web, and soon from our app, email and Telegram.
Call it from your stack
curl -X POST https://api.kit.forhosting.com/audio/meeting-notes \
-H "Authorization: Bearer $KIT_KEY" \
-H "Content-Type: application/json" \
-d '{"audio":"https://ejemplo.com/audio.mp3"}'const res = await fetch("https://api.kit.forhosting.com/audio/meeting-notes", {
method: "POST",
headers: {
"Authorization": `Bearer ${process.env.KIT_KEY}`,
"Content-Type": "application/json"
},
body: JSON.stringify({
"audio": "https://ejemplo.com/audio.mp3"
})
});
const { task_id } = await res.json();import os, requests
res = requests.post(
"https://api.kit.forhosting.com/audio/meeting-notes",
headers={"Authorization": f"Bearer {os.environ['KIT_KEY']}"},
json={
"audio": "https://ejemplo.com/audio.mp3"
},
)
task_id = res.json()["task_id"]<?php
$res = file_get_contents("https://api.kit.forhosting.com/audio/meeting-notes", false, stream_context_create([
"http" => [
"method" => "POST",
"header" => "Authorization: Bearer " . getenv("KIT_KEY") . "\r\nContent-Type: application/json",
"content" => '{"audio":"https://ejemplo.com/audio.mp3"}',
],
]));
$task = json_decode($res, true);body := bytes.NewBufferString(`{"audio":"https://ejemplo.com/audio.mp3"}`)
req, _ := http.NewRequest("POST", "https://api.kit.forhosting.com/audio/meeting-notes", body)
req.Header.Set("Authorization", "Bearer "+os.Getenv("KIT_KEY"))
req.Header.Set("Content-Type", "application/json")
res, _ := http.DefaultClient.Do(req)Example request
{
"audio": "https://ejemplo.com/audio.mp3"
}Example response
{
"task_id": "tsk_a1b2c3d4e5f6a1b2c3d4e5f6",
"type": "audio.meeting_notes",
"status": "queued",
"_links": {
"result": "/tasks/tsk_…/result"
}
}The API is asynchronous: the call returns a task_id immediately and the result arrives by webhook. Polling is capped at 1 req/s per task.
Pricing
Published price — no tokens, no invented credits. A failed task is never charged.
Limits
max_mb | 200 |
max_minutes | 180 |
Errors
| HTTP | Code | Meaning |
|---|---|---|
401 | unauthorized | Missing or invalid API key. |
402 | insufficient_balance | Your balance doesn't cover the task price. |
404 | unknown_type | That task type doesn't exist. |
429 | rate_limited | Too many requests. Use the webhook instead of polling. |
422 | task_failed | The task failed after 3 retries. You are never charged for it. |