Help arrives at your cursor
Select text anywhere you read or write and the menu comes to you.
Informed by the full picture
Every suggestion fits the thread, the document, or the page your selection sits in.
Changes only what you select
Revisions touch the text you selected and leave the tone and structure around it intact.
Skip the round-trip
Getting help usually means leaving your work: Copy the text, explain what it's part of, paste the answer back, then make it fit again. Inline intelligence acts on the text in front of you, so the answer arrives already shaped by what surrounds it.
Informed by the full picture
Inline intelligence already has the context of your window, so the responses are relevant to doc, article or email you're working on.
Edits that still sound like you
Rewrites refine exactly what you pointed at, in the voice the rest of your writing already established.
Works across email, docs, and the web
Inline intelligence works in Gmail, Outlook, Slack, Google Docs, Notion, and the web. It appears only when you select something, and you can snooze it or switch it off at any time.
What inline intelligence can help you with
Writing
Rephrase
Say it another way, in your own voice
Shorten
Tighten it without losing the point
Proofread
Catch errors without a separate pass
Reading
Explain
Understand any term or passage in context
Translate
Work across languages without leaving the page
Scan for AI
Check whether a passage was likely AI-generated
Going further
Ask Go
Follow up on anything
Open the Go side panel
For longer or deeper work
Contextual prompts
Suggested actions based on what you’re doing
Why use inline intelligence
Nothing to set up
No install, no configuration. Select text and it's there.
Help before you ask
The menu offers actions that fit what you selected.
Reading and writing
Inline intelligence handles both from the same selection.
Proofread in place
Spelling, punctuation, and grammar fixed in place, with no extra edits.
How to use inline intelligence
Step 1: Select any text.
In any document, email, or web page, select text and click the Go icon.
Step 2: Pick an action.
A menu appears with options like Proofread, Translate, Explain, and Scan for AI.
Step 3: Keep it or go further.
Accept the change in place, ask a follow-up in context, or open the Go side panel for deeper work.
Frequently asked questions
Inline intelligence is a context-aware AI layer that lives at your cursor inside Superhuman Go. Select text in a document, an email, or a web page and it suggests the most relevant actions for what you selected, using the thread, document, or page around it as context.
Chat tools only know what you paste into them. You go to them, paste your text, explain the surrounding context, and get back a full response that you then reconcile with what you had. Inline intelligence works inside the surface you're already in, sees what surrounds your selection, and changes only what you asked it to.
Rephrase, Shorten, Proofread, Explain, Translate, Scan for AI, and Ask Go. Additional actions surface based on what you selected and where, so the most relevant options are always available.
Gmail, Outlook, Slack, Google Docs, Notion, and web pages. Inline intelligence follows you without asking you to change where you work.
No. Inline intelligence is available from any text selection once you have Superhuman Go.
Yes. Inline intelligence is included on the Superhuman Go free plan. View our plans page for more information.
No. Edits apply to your selection and preserve the tone and structure of the text around it. You can preview the change in place, retry it, or revert it, and nothing is committed until you accept it.
It reads what you selected, along with the surrounding material (the document, the thread, or the page), then offers the actions most likely to help with that kind of content.
Yes. Snooze inline intelligence when you need uninterrupted focus, or turn it off entirely. It appears when you select text and disappears when you're done, with no persistent panel to dismiss.
Knowledge Checker will be available from the inline intelligence menu, so you'll be able to check a claim from the same selection.
Inline intelligence follows Superhuman Go's data practices. We don't train third-party models on your data.
For Superhuman's own models, Go respects your Customer Data Training Control settings. Access to your connected sources is read-only, and inline intelligence makes no new use of your connectors beyond the terms you agreed to when you first authorized them.
Learn more about Superhuman Go's data and privacy policies.
