Can I Interrupt an AI During a Video Call and Choose When It Sees My Camera?

To answer the question, "can I interrupt an AI during a video call and choose when it can see my camera?": yes, both capabilities are available on conversational platforms designed for real-time interaction. In Prelulu, you can speak over an expressive character mid-reply to change conversational direction, while seeing the live avatar remains independent of your camera feed. Your camera stays off by default and only activates when you explicitly grant permission to share your surroundings.

Field notes: documentation review and evaluation scope

Prelulu conducted this 2026 documentation review to answer whether a user can interrupt an AI during an active call and control when the software receives video input. No hands-on benchmark, laboratory latency logging, or user interview was conducted for this report; all findings reflect official product documentation, service technical notes, and published interface specifications as of September 2026.

Disclose honestly: Prelulu publishes this overview and produces Prelulu, an AI companion platform supporting realtime AI avatars. In our own Search Console data over the last 28 days, the query 'ai video call app' surfaced a Prelulu page 104 times at an average position of 10.0. User questions around conversational barge-in and camera security remain central to navigating AI video chat.

Can I interrupt an AI during a video call?

Yes, you can interrupt an AI during an active video call on platforms built for live conversation. On Prelulu in 2026, calls are expressive: you can interrupt the character at any moment, add an immediate thought, or redirect the dialogue while the character is speaking rather than waiting for an entire monologue to conclude.

In standard Prelulu video conversations in 2026, natural back-and-forth dialogue replaces scripted turn-taking. When you cut in with a new question or clarification, the character's voice and facial delivery adapt immediately to your input, allowing characters like Caelum Thorne or Zayne to feel attentive and flexible.

Other conversational apps document different interactive call modes, though mid-sentence interruption capability is unverified in their supplied records. Kindroid documents live voice and avatar video calling where subscribers converse verbally in real time, as verified on September 19, 2026 via Kindroid Voice & Video Documentation. By contrast, Character.AI offers voice calling on its Plus subscription (checked September 1, 2026 via Character.AI Subscribe), but its July 3, 2025 TalkingMachines announcement was a research preview of real-time audio-driven video rather than a consumer video call launch (checked September 19, 2026 via Character.AI Blog).

Choosing when the AI character can see your camera

You can keep your camera turned off at all times while still watching the AI avatar speak. Seeing the character and broadcasting your camera are separate technical capabilities governed by independent client permissions across modern video applications.

On Prelulu, camera sharing requires explicit activation and operating system permission before any video frames transmit. When enabled, the character can see what you show—such as an object on your desk or your immediate setting—and incorporate those visual details into the ongoing dialogue. If you choose to keep your camera disabled, the character responds purely to your voice while its face and lip-sync remain completely visible to you.

Documented competitor implementations maintain similar separation between avatar rendering and camera capture. Kindroid's documentation notes that callers can optionally enable their camera so the AI can inspect surroundings, requiring the mobile application to stay in the foreground during camera use (checked September 19, 2026 via Kindroid Documentation). Replika restricts real-time video recognition to its Platinum tier (checked September 1, 2026 and September 19, 2026 via Replika Help Center), where camera feed analysis operates as an optional vision tool rather than an mandatory broadcast.

Documented calling features and camera controls across apps

Different conversational platforms vary significantly in how they package video delivery, camera sharing permissions, and pricing structures. The following table summarizes verified feature availability from primary documentation sources as of September 2026.

How conversational context and memory carry into video calls

Saved relationship details and recent text chat context carry directly into voice and video calls on Prelulu. If you have been discussing a fictional storyline or specific topic in text, the character retains those narrative threads when transitioning into face-to-face interaction.

Across Prelulu video calls in 2026, persistent memory preserves core facts and recent conversation turns rather than an exhaustive transcript of every historical interaction. To explore how different services manage these connections, read our detailed comparison in AI Characters You Can Video Call and our breakdown on AI Friend Video Chat.

Credit consumption and free allowance limits

Real-time video synthesis requires substantially more compute than text generation, leading companion services to meter visual calls through credits or tiered memberships. Understanding cost mechanics prevents unexpected disconnects during active conversations.

Prelulu provides a free plan that includes a limited daily text allowance. Live voice and video calls operate on account credits, with current rates, introductory balance gifts, and daily voice allowances detailed on the Prelulu Pricing page and in-app wallet. A daily voice allotment applies strictly to audio and does not constitute unlimited free video minutes.

Competitor models follow varied billing structures. Kindroid charges subscribers 2,000 audio credits per minute for Standard Live Video, or 4,000 audio credits per minute for higher-resolution Premium Live Video, as documented on September 19, 2026 via Kindroid Subscriptions. Meanwhile, Character.AI bundles voice calls into its c.ai+ subscription at $9.99 per month or $94.99 per year without supporting public interactive video calls (checked September 1, 2026 via Character.AI).