Feature request: live vision (screen sharing + camera) in Claude voice mode

Status Open
Maintainer reply None cached
Activity 0 comments · opened Aug 6, 2026

I'm an Advanced Care Paramedic who uses Claude as a full personal assistant — on my Mac through Claude Code and heavily by voice on my phone. This is a follow-up to my earlier voice request (#83434): the next gap after persistent voice is vision during voice.

The request: let Claude see my screen (and optionally my camera) while we talk in voice mode.

Today voice mode is audio-only. The workflow I keep wanting is simple: I'm working through something online — navigating a website, filling out a form, comparing options, learning new software — and I want to share my screen and just talk, with Claude watching what I'm doing and explaining things in real time as I go. "What does this setting do?" "Which of these should I pick?" "Where do I click next?" — answered live, in context, without me stopping to screenshot, upload, and type a description of what I'm looking at.

That screenshot-and-describe loop is the current workaround, and it breaks the flow completely. Live screen sharing in voice mode would turn Claude into a real over-the-shoulder guide.

The competitive gap is real: ChatGPT's voice mode, Google's Gemini Live, and Grok all support live camera and/or screen sharing in their conversational modes. Claude has the best reasoning to bring to what's on my screen — it's the one notable assistant that can't currently look.

For working professionals this is the difference between an assistant you consult and an assistant that works alongside you. Combined with persistent voice sessions (#83434), screen-aware voice conversation is the full "assistant on a call with you" experience.

View original on GitHub ↗