Upgrading to human transcription

By default, Reduct generates a computer-generated transcript for every recording you upload or import.

These automatic transcripts are fast and fairly accurate, making them great for quick reviews and edits.

If you need a higher level of accuracy, especially for complex audio with crosstalk, strong accents, or technical terminology, you can upgrade to human transcription.

Our team of professional transcribers typically delivers the transcript overnight.

Setting the default transcription setting to human transcription

1. You can change the default transcription setting of a project to Human transcription. This will ensure that all the media that you upload inside that project will be automatically upgraded to human transcription.
2. Click on the three-dot kebab (︙)menu next to your project.
3. From the dropdown menu, select Transcription settings.
4. In the pop-up window, click on the Transcription Method and select Human.
5. You can also add a Glossary to add preferred spellings, abbreviations, names, or any custom requirements that you’d like to communicate to the human transcribers.
6. Then, click on the Update button.

This will change your entire project’s setting to human transcription and all the recordings that you will import within that project will be upgraded to human transcription moving forward.

💡Transcription glossary and hints

The transcription glossary is useful for listing expected names of people, technical terms, and jargon that are mentioned or likely to come up in your recordings in the project.

When you provide a glossary, our transcribers use it as a reference to ensure consistent and accurate transcription. This helps reduce errors, especially for content with domain-specific language or uncommon names.

Upgrading a specific recording to human transcription

There may be cases where you’d want to upgrade a specific AI transcription on a case-by-case basis. For that, you can keep the default transcription setting of the project to Computer.

First, open the recording you want to upgrade.

1. Then, on the top of your recording’s transcription, you’ll see the Computer transcript button,click on it.
2. Now, you’ll see a pop-up window with an option to upgrade to the human transcript, click on the Upgrade transcript button.

Once your human transcription is ready, you'll receive an email notification.

The upgraded transcript will automatically replace the computer-generated version with all your existing highlights and tags intact.

FAQs:

How long does human transcription take?

Human transcripts are typically delivered overnight. Depending on the length and quality of your audio, turnaround time ranges from 6 to 24 hours.

Can I continue working while the human transcription is in progress?

Yes. You can continue working with the computer-generated transcript while your human transcript is being prepared.

Can I change speaker names before or after the upgrade?

While you can highlight, tag, and make comments to the AI transcript, we advise not to make corrections to the transcripts itself until the human transcription is ready.

Therefore, only change the speaker names after you receive the upgraded human transcription.

How much does it cost to upgrade to human transcription?

Our human transcription is priced at $1.25/minute of audio. Reduct automatically detects extended silences from your recordings ensuring you're only billed for the actual spoken content and not for long pauses or silent sections.

How accurate are Reduct transcripts?

We regularly benchmark our transcription accuracy to make sure we are one of the best transcription providers in the word. You can see the Transcription Benchmark Report 2025here→

Generally, the human transcription is usually ~99% accurate in both word and speaker accuracy, while automated transcripts are ~94% accurate, for both word and speaker accuracy.