Demo Audio Looks Clean? Choose for Your Audio First
AgentLeverage Team
4/15/2026
Vendor demo audio is a quiet room and one native speaker. Your file is a laptop mic, two people talking over each other, and a last name the model has never seen.
The clip looks clean. That is not your recording.
Choose transcription software against the audio you actually have. Test your audio first. You still listen back.
Why vendor demos lie
Every vendor leads with an accuracy number. That number is almost always measured on clean audio: one speaker, a quiet room, a studio mic.
Your recording is usually none of those:
- Accents and regional speech the demo never used
- Legal, medical, or product jargon
- People talking over each other
- Phone audio, room echo, or a cheap laptop mic
- Two languages in one conversation
A clean demo does not tell you what happens to your file. Word-error arithmetic on a vendor percent is theater. It assumes the percent applies to your recording. It does not.
The honest check is to run your own audio and listen back. Wrong labels and dropped turns show up there, not on the marketing page.
What label-speakers-first actually returns
Speaker Separation takes a recording and returns a labeled transcript.
Upload the file. The run labels each voice (Speaker 1, Speaker 2). You rename those labels. You can download isolated stems (one audio file per voice) and export txt or srt.
Guest runs are free (2/day). Signed-in orgs pay 2 credits per minute of audio, rounded up, with a 1-minute floor.
That is the job: who said what on a recording you already have. Observed use here is people labeling speakers. It is not a certified transcript, and it is not a meeting bot that joins Zoom.
You still review. Overlaps and wrong labels still need a listen.
If the file is any multi-speaker recording and you still need who said what: how to split audio by speaker, then rename the labels
Other tool categories
These are other products. We do not publish their prices or accuracy percents. Pick the category that matches the job, then test on your audio.
Meeting bots (Otter, Fireflies, Fathom, tl;dv)
A bot joins Zoom, Teams, or Meet and transcribes the call. Useful when the job is "catch the meeting from the calendar." Not useful for a field recording, a phone export, or a file you already have.
We do not auto-join Zoom, Meet, or Teams.
Upload tools (Rev, Sonix, Descript)
You upload an audio or video file. Useful when the recording is not a video call: interviews, customer calls, podcasts. Descript is built around editing the recording from the transcript. Evaluate that workflow if editing is the job.
APIs (Deepgram, AssemblyAI, OpenAI Whisper)
Raw transcription for a developer building a pipeline. Useful when you want custom vocabulary or batch jobs in your own app. Not a button for an owner who just has a file.
Human + AI hybrid (Rev Human, Scribie)
A person reviews the machine draft. Useful when you need a certified or official copy: depositions, medical records, anything you will file.
We do not sell that. Speaker Separation is a working labeled transcript. You still review.
Weekly team meetings. A meeting bot. The calendar join is the product.
A file you already have. Upload. That is our lane: Speaker Separation.
A certified copy. Human hybrid. Not us.
Building it into an app. An API.
If the export already says Speaker 1 for an hour: label who spoke on a Zoom, Meet, or Teams recording
What we do not do
Speaker Separation does not:
- Return a certified transcript or a court record
- Auto-join Zoom, Meet, or Teams
- Invent real names (you still rename Speaker 1)
- Beat Otter on price. We do not claim that.
- Send a recording into Meeting Summarizer as one run
Meeting Summarizer does not transcribe audio. It does not take a recording. Transcribe and recap are two tools. If you want a recap, you already need a transcript you checked.
We do not train models on your files. Start with 30 free credits. Pack credits never expire. Unused included credits do not carry forward. Paid checkout charges immediately. There is no 7-day trial.
If the job is a deposition recording and you still wait on the certified copy: label speakers for a working transcript, not a court record
Optional recap after you check the transcript
Once you have a transcript you have already checked, Meeting Summarizer can turn pasted notes or that transcript into a recap, decisions, owners, and action items when those were said.
Guest runs are free (5/day). Signed-in orgs pay 2 credits per summary. It accepts TXT, DOCX, VTT, and SRT. Meeting Summarizer does not transcribe audio.
Skip it if you only need the labeled transcript.
How to test on your audio
Pick a real recording. Not the vendor clip.
- Upload it to Speaker Separation. Guest runs are free (2/day).
- Rename Speaker 1 and Speaker 2.
- Listen back. Check overlaps and wrong labels. Download stems or export txt/srt if you want them.
- Optional: paste the transcript you checked into Meeting Summarizer for a recap.
- You still review before you share.
No account for a guest run. Sign in if you want the job kept. Start with 30 free credits. Pack credits never expire. Unused included credits do not carry forward. Paid checkout charges immediately.
Frequently asked questions
Does this replace a meeting bot like Otter?
No. Otter and Fireflies join the call. We label speakers on a file you upload. We do not auto-join Zoom, and we do not claim to beat Otter on price.
Can I try it without an account?
Yes. Guest runs are free (2/day). Sign in if you want the job kept until you delete it.
What does a signed-in run cost?
2 credits per minute of audio, rounded up, with a 1-minute floor. Start with 30 free credits. Pack credits never expire. Unused included credits do not carry forward. Paid checkout charges immediately. There is no 7-day trial.
Does Meeting Summarizer transcribe audio?
No. Meeting Summarizer does not transcribe audio. Paste notes or a transcript you already checked (TXT, DOCX, VTT, SRT). It does not take a recording.
Is this a certified transcript?
No. It is a working labeled transcript. You still review. Do not file it as an official record.
Need who said what on a recording that is not the vendor demo? Upload one. Guest runs are free (2/day). You still listen back.
Try it yourself
Speaker Separation
Upload a recording. This speaker separation labels multi-speaker audio. Rename Speaker 1 and download isolated stems plus a timestamped transcript.
Start with 30 free credits — no card required.
Related Articles
Vendor Says 95%? Verify Who Said What First
Vendor percents are clean-room theater. Verify on your audio first. Guest 2/day. You still listen back.
Still Untangling Who Spoke? Audio Diarization Tool
Still reading a nameless wall of text? An audio diarization tool labels who spoke. You still rename Speaker 1 and review. Guest 2/day.
Speaker Diarization with n8n: Label Who Said What on a Call
Upload a call recording to an n8n form and download a .txt transcript with a name on every turn. Free template and the n8n-nodes-agentleverage node.
Get weekly AI tips
Practical AI productivity tips every week. No fluff.