Short answer: open a request in Postman, click the Docs tab or a Params, Headers, or Examples field, press ⌃+⌥+R (Mac) or Ctrl+Alt+R (Windows/Linux), talk through the endpoint, press the hotkey again. AICHE inserts text into that field. Mock servers and the public API portal at <workspace>.postman.co render the same descriptions you fill in.
Postman surfaces documentation in more places than most teams use: Collection Overview, folder descriptions, the request Docs tab, Params and Headers description cells, saved Examples, and notes on mock-server branches. When a GET /customers/{id} request finally returns 200, the fields that should explain path variables, auth, pagination, and error shapes are usually still empty.
Concrete example: click the Docs tab on GET /customers/{id}, press the hotkey, and say: "Returns one customer by ID. Requires Bearer token scoped to customers:read. Path variable id is the UUID from the list endpoint. Response includes created_at and nested address. Returns 404 if deleted or outside the caller's org. Paginated list lives on GET /customers with page and limit query params." Thirty seconds of speech fills the portal page a reviewer actually reads.
The Problem
Almost none of those fields get written. You are tuning auth headers, watching a 401 turn into a 200, and copying a JSON response into an Example. Stopping to type the pagination note breaks your flow, so the API portal stays thin.
What Changes
Right after a request turns green, you stay in Postman, click into the description you are already staring at, and talk for 20 or 30 seconds. What does it do, who calls it, what auth it expects, what comes back, what error cases actually matter. Press the hotkey, speak, press again. AICHE fills the text field for you, and you move to the next request.
Math: speaking sits around 150 words per minute. Typing sits around 40. A 100 word endpoint description that takes 2 to 3 minutes to type takes about 40 seconds to say out loud, and the spoken version usually includes the edge cases you would have skipped if you were typing.
How It Works With Postman
- Open the Postman desktop app on macOS, Windows, or Linux.
- Select a request from your collection in the sidebar.
- Click the Docs tab, the description box under the request name, a parameter description cell, or an example notes area.
- Put the cursor where you want text.
- Press ⌃+⌥+R (Mac) or Ctrl+Alt+R (Windows/Linux) to start recording.
- Speak your description in one clear pass.
- Press the hotkey again to stop. AICHE transcribes and inserts text into the focused field.
- Add code fences, tables, or links inside Postman if needed.
AICHE never sends requests, runs scripts, or clicks buttons. Audio is streamed for cloud transcription, processed, and discarded within 1 second. No persistent audio copy.
The Overview tab and Docs tab both accept Markdown. AICHE handles the narrative sentences. You apply headings, code blocks, and tables once the text is in place.
Where Dictation Belongs In Collections
Postman has more description fields than most collections use. AICHE helps you actually fill them.
Collection Overview Tab
This is the front door of your API. Use voice to introduce what the API does, which environments exist, how auth works, what the base URL is, and what the standard error shape looks like. Then paste a small JSON error example in a fenced code block.
Folder Descriptions
Folders like Auth, Users, Orders, or Webhooks all have their own description. Dictate the shared rules for each group: scopes, base paths, and typical call sequences. That replaces repeated boilerplate on every individual request.
Request Docs Tab
The Docs tab is the main story for each endpoint. Dictate what the operation does, when clients should call it, which parameters are required, how pagination works, and which status codes you actually emit. This is the content that becomes the core of the public docs page for that request when you publish.
Parameter, Header, And Path Variable Descriptions
Every row in Params, Headers, and Path Variables has its own description column. This is where you explain allowed values, defaults, units, and failure behavior. Click the cell, hit the hotkey, say two sentences, stop the recording, then move down the table. It is tedious to type and easy to speak.
Example Notes
Each saved Example in Postman (the request and response pair) has a description field. Dictate which scenario a given Example represents, like "valid request with default pagination", "missing Authorization header", or "rate limit exceeded with retry-after header". When you run a mock server from the collection, those descriptions are what mock consumers read to understand which branch they hit.
API Spec Summaries
If you maintain an OpenAPI or AsyncAPI spec in the APIs section, you can dictate summaries and descriptions directly into the operation editor. AICHE still just types text. Postman compiles that spec into the reference and your public docs stay in sync.
Mock Servers And Public Docs
Postman mock servers match incoming traffic to Examples mechanically, but the story behind each example is up to you. For every success, validation error, partial result, or rate limit Example, open the description field and speak:
- What the client sent
- Why this response shape exists
- What a caller should do next
Those notes carry through to the mock server UI and the public docs portal at <workspace-name>.postman.co/.... There is no second docs site to maintain. Dictate while the scenario is fresh, then publish from Postman.
Flows, Scripts, And Modern API Workflows
Postman Flows and the unified Scripts editor are where a lot of modern API logic lives. Flows orchestrate branching requests. Scripts sign requests, manage tokens, and set environment variables. Newer AI features help you reason about responses and collections, but the configuration still lives as text in front of you.
Use AICHE to explain that logic the way you would explain it on a call:
- In the Scripts pane, place the cursor above a block and dictate a short comment describing what it does and why.
- In Flows, put the cursor in a node note and describe what inputs it expects and what outputs it produces.
- In prompts you send to Postman AI or similar tools, dictate a tight natural language query, then review and edit it before you press the Ask or Run button.
AICHE never executes Flows, runs scripts, or sends any request. It only writes the words that describe what you are about to run or what you want an AI assistant to do.
Environments, Variables, And Collections At Scale
Postman environments like Local, Dev, Staging, and Production hold variables that control how your requests behave. Each variable has a description slot. Open the environment editor, click the description cell, and dictate what the variable is, where its value comes from, and what breaks if it is wrong. New teammates can then understand an environment without a separate onboarding doc.
At collection scale, the same pattern holds:
- Dictate once into collection and folder descriptions so the big picture is clear.
- Dictate into individual request Docs tabs for detailed behavior.
- Dictate into parameter and header rows where precision matters most.
Postman turns those descriptions into internal documentation and, when you publish, into your external portal.
Speaking In Code Heavy Contexts
API docs are full of precise identifiers like Authorization, Bearer, application/json, retry-after, created_at, and idempotency-key. Turn on AICHE's Software Development profile so the recognizer is tuned for code shaped speech, including HTTP codes and header names. Add your service names, internal acronyms, and important field names to custom vocabulary so they are spelled your way every time.
Privacy, Platforms, And Limits
- Postman desktop workflow: AICHE's desktop hotkey types into whichever Postman text field has focus on macOS, Windows, or Linux. AICHE also offers mobile, Chrome, and other surfaces; this page targets the Postman desktop client in a browser or app window.
- Zero-retention audio: audio is streamed for cloud transcription, processed, and discarded immediately after processing, within 1 second. No persistent audio copy.
- You stay in control: AICHE does not click Send, call your APIs, change environments, or enable scripts. It only fills text you can see and edit before you commit or publish.
Common Questions
Q: Does AICHE work in the Postman desktop app or the web workspace?
A: The desktop client is the recommended setup. AICHE is a desktop app and types into the focused text field on your machine. If you open Postman in a browser tab, it can still insert text there, but the native Postman desktop app gives the cleanest experience.
Q: Will it format JSON or write tests for me?
A: No. AICHE turns speech into prose. The workflow is to dictate the explanation, then paste JSON or write tests manually in Postman. That keeps code and assertions explicit and reviewable.
Q: My API names are strange. Will transcription break them?
A: You can add custom vocabulary entries for endpoint names, service names, and internal acronyms. Once added, AICHE spells them consistently every time.
Q: Can I dictate directly into a Postman collection file in git?
A: Yes. If you open the v3 collection YAML in your editor, AICHE will type wherever your cursor is. Most teams prefer dictating inside Postman itself and letting Postman handle the serialization.
Q: Does this change what appears in the public portal or only in the workspace?
A: Both. The public docs portal renders the same description fields you filled in the Overview, Docs tabs, parameter rows, and Example notes.
Result: the collection that used to have blank descriptions and thin public docs gradually fills in as you work. You keep building and testing in Postman, but the docs layer grows at speaking speed instead of typing speed.
Try it now: open a collection you use every day, click into the Overview tab, press your hotkey, and spend 30 seconds verbally explaining what the API is and how auth works. That one paragraph becomes the front door of your Postman docs.