Accessibility

Read, listen, ask.

Spoken help with text and the camera view, captions for conversations, and ways to use Avenkin by voice. Start on your iPhone, with headphones or AirPods, or with compatible glasses.

Current capabilities

Help with everyday information.

Reading Assistant and Blind Assistant are available live presets. Features depend on the selected model, permissions and camera view.

Read

Hear text aloud

Ask for text from a label, menu, sign or document. Reading Assistant focuses on visible text and can help translate or summarise it.

“Read this label.”

Describe

Ask about the camera view

With a vision-capable model, ask for a spoken description or follow up about a detail. The answer depends on what the camera can capture.

“Describe what is in front of me.”

Follow along

Captions and translation

Enable ambient captions for a running transcript, or translate spoken language. Language support and offline availability depend on your speech and translation setup.

“Start translating Spanish to English.”

Start

Use voice and shortcuts

Start requests with the microphone control, an enabled wake word or Siri. App Shortcuts can also be assigned to the iPhone Action button.

“Hey Siri, ask Avenkin a question.”

Operate

VoiceOver and spoken feedback

The app includes VoiceOver labels, spoken feedback and session announcements. Complete setup and recovery with blind participants are still being validated.

Listen

Choose your audio route

Use your iPhone or iPad, headphones, AirPods or compatible glasses for spoken replies. Check both the microphone and where replies play before starting a longer session.

First use

Start with one useful task.

  1. Start with the beta on an iPhone running iOS 26 or later. You can begin with the device camera; glasses are optional. The standard phone version is also usable on iPad and iPhone Duo; dedicated versions are coming soon.
  2. Choose your services. Set up an AI model, speech recognition and a spoken voice. Descriptions need a vision-capable model.
  3. Grant the permissions the task needs. Check camera and microphone access, then ask a short question to confirm the audio route.
  4. Try Reading Assistant or Blind Assistant. Begin with clearly printed text and a fresh camera capture. Ask a follow-up to check the result.

Compatible Meta glasses need their own pairing, registration and camera permissions. The first-use guide covers these steps, and the accessibility notes describe current limits.

Your setup, your choice

Know where the processing happens.

No separate accessibility unlock

Accessibility features have no separate paid unlock. Your selected cloud AI or speech provider may charge for usage.

Local storage, chosen services

Saved context is stored on your device. Images, audio or recalled context may go to the services you select. Read the privacy policy.

Prepare for offline use

Use Apple Intelligence where available or a compatible downloaded model, with local speech and voice. Apple Intelligence supports text conversations and selected app tools in Avenkin; visual descriptions need a vision-capable model. Automatic offline takeover and reliable local vision with a pocketed phone are not established.

Work in progress

Built features need real-world validation.

Independent setup, reliable reading and recovery during daily use still need testing with blind participants and real hardware. Avenkin can misread text or miss details; it supplements existing mobility aids and cannot establish that a route or crossing is safe.

We want to progress this with the people who will use it. Feedback on setup, spoken feedback, reading and recovery will help set the priorities.