Speech-to-text (AI transcription)
Definition
Speech-to-text is technology that turns spoken audio into written text. AI transcription models do this for recordings such as meetings and calls, so the words can then be searched, summarized or turned into tasks.
By Ronan Liedmeier, fractional head of growth and course author · Published 17 October 2026 · Last updated 17 October 2026
Also called: AI transcription, automatic speech recognition (ASR)
For work, it makes recorded talk usable: a client call becomes notes you can search, and a voice memo recorded on a walk becomes a first draft.
How it works
A transcription model takes an audio file or a live stream and returns text, often with timestamps. Some services can also label who is speaking, known as speaker diarization. OpenAI's speech-to-text API, for example, has a separate model for speaker labels and a separate endpoint that translates speech into English text.
Dictation and voice mode both use speech-to-text but do different jobs. Dictation types what you say. Voice mode is a spoken back-and-forth with an assistant, useful for thinking out loud.
In a workflow, transcription is the first step: record, transcribe, extract decisions and action items as structured output, then send them where they belong.
Example
After each weekly client call, an account manager's recording is transcribed automatically. A model pulls out decisions, owners and dates, then drafts a recap email and a CRM note. She reviews both before anything is saved or sent, and on one call changes "budget approved" to "budget likely, waiting on CFO sign-off".
Common mistake
Recording people without asking. Consent rules differ between countries and states, so ask at the start of every call whether you can record and transcribe it. Make the request part of the meeting routine so nobody has to remember it.
Learn it in AI Academy
- Voice & AudioBeginner · 9 min
- Voice & Image in WorkflowsAdvanced · 13 min
- Meeting Intelligence StackAdvanced · 15 min
Related terms
- Multimodal AI
- Multimodal AI is AI that can work with more than one kind of input or output, such as text, images, audio and video, in the same conversation. You can show it a photo or screenshot and ask about it, or describe a picture and get one made.
- AI workflow
- An AI workflow is a fixed series of steps, set in advance by a person or by code, in which an AI model handles one or more of the steps. The steps run in the same order every time, which makes a workflow easy to test and check.
- Structured output (JSON mode)
- Structured output is a way of getting an AI model to return its answer in a fixed, machine-readable format, usually JSON with named fields, instead of free text. It lets another program read the answer reliably, for example to fill a spreadsheet row or pick the next step in a workflow.
- Human in the loop
- Human in the loop is a setup where an AI system pauses at set points so a person can approve or correct its work before it takes effect. It is used for actions that are hard to undo, such as sending emails to customers or moving money.