Be My Eyes

Overview and architecture

Be My Eyes is a global assistive technology platform designed to provide visual assistance to blind and low-vision individuals. Founded in 2012 by Danish furniture craftsman Hans Jorgen Wiberg and headquartered in Copenhagen and San Francisco, the platform originally established a crowdsourced video network pairing visually impaired users with sighted human volunteers. In early 2023, Be My Eyes expanded its technical architecture through a technical partnership with OpenAI, deploying Be My AI, an autonomous visual interpreter powered by GPT-4 with Vision (GPT-4V) and multimodal language models.

The system architecture integrates computer vision processing, real-time WebRTC audio-video streaming, and multimodal large language model inference. When a user captures a photograph or directs a camera feed within the mobile application, the image routes securely to OpenAI vision inference clusters. The underlying model analyzes scene geometry, identifies text through optical character recognition (OCR), differentiates colors, measures relative spatial distances, and synthesizes a descriptive natural-language summary. Users interact with the AI assistant conversationally through voice or screen readers, asking contextual follow-up questions such as ingredient verifications, clothing coordination, or appliance settings.

A central engineering strength of Be My Eyes is its dual-path redundancy. When autonomous AI descriptions require human confirmation, or when a user encounters complex safety-critical navigation scenarios, the application switches directly to live human assistance. Users can connect within thirty seconds to a verified sighted volunteer from a network of over eight million participants speaking more than 185 languages, or route their call to dedicated corporate representatives via the Specialized Help service directory. The platform is accessible globally on iOS and Android smartphones, alongside wearable implementations on smart glasses hardware.

Key capabilities and functional modules

Be My Eyes organizes its visual assistance tools into specialized functional modules serving daily personal independence and corporate accessibility needs.

• Be My AI autonomous visual interpreter: Generates conversational scene descriptions from smartphone photographs, recognizing objects, physical layouts, facial expressions, and complex visual contexts.
• Conversational context and follow-up dialogue: Allows users to ask detailed follow-up questions about captured images via voice synthesis or screen reader input without recapturing photographs.
• Instant text reading and document transcription: Transcribes printed street signs, prescription bottle labels, digital monitor displays, and paper correspondence with high character accuracy.
• Crowdsourced live video network: Connects visually impaired callers with vetted sighted volunteers through one-way video and two-way audio streaming in over 185 languages.
• Specialized Help enterprise directory: Connects users directly to trained customer service agents at partner organizations, including Microsoft, Google, Sony, Hilton, and Procter and Gamble.
• Smart glasses and wearable integrations: Interfaces with smart eyewear platforms such as Ray-Ban Meta glasses and Envision glasses, providing hands-free visual analysis.
• Color and pattern identification: Distinguishes subtle clothing shades, wall paint swatches, and fabric patterns to assist in wardrobe coordination and home organization.
• Barcode and product label identification: Identifies packaged consumer goods, expiration dates, cooking directions, and nutritional data tables in grocery environments.
• Indoor scene exploration: Describes room layouts, furniture obstacles, doorway locations, and elevator button panels to assist indoor navigation.
• Complete screen reader accessibility: Operates in full compliance with Apple VoiceOver and Android TalkBack screen reader interfaces, utilizing haptic and audio navigation cues.

Competitor comparison

The assistive technology sector includes several specialized computer vision tools. The table below compares Be My Eyes against Microsoft Seeing AI and Envision across key operational dimensions.

Practical use cases and target audiences

Be My Eyes serves a diverse population of blind, low-vision, and visually impaired individuals while assisting enterprise customer support teams.

• Daily household management: Users read expiration dates on milk cartons, differentiate canned pantry goods, identify oven dial settings, and check refrigerator contents without assistance from family members.
• Personal wardrobe and grooming: Individuals check clothing items for stains, confirm matching sock pairs, verify makeup shades, and confirm shoe colors before departing for professional meetings.
• Transit and urban wayfinding: Commuters read airport departure gate monitors, confirm bus route numbers, identify street signs, and locate elevator call buttons in unfamiliar office buildings.
• Technical device setup and troubleshooting: Blind tech users point their camera at computer monitors displaying BIOS errors, router lights, or television input menus, receiving actionable descriptions to resolve hardware issues.
• Retail shopping and packaging: Shoppers navigate supermarket aisles, compare clothing price tags, read ingredient warnings on cosmetics, and inspect electronics packaging independently.
• Enterprise customer service accessibility: Major corporations employ the Specialized Help directory to provide accessible customer onboarding, technical troubleshooting, and product support to blind consumers.

Hardware, quota and technical limitations

While Be My Eyes delivers high real-world utility, users should recognize specific technical parameters and operational constraints.

• Network connectivity requirement: Both Be My AI multimodal queries and live volunteer video calls require an active internet connection (4G, 5G, or stable Wi-Fi). The software does not execute full offline multimodal inference on device hardware.
• Camera autofocus and lighting dependency: In poor lighting conditions or with blurred camera focus, vision models can misidentify dense text or small visual details. Adequate ambient illumination improves recognition accuracy.
• Safety and navigation disclaimers: The service is explicitly not certified as a primary mobility aid or obstacle avoidance sensor. Users must rely on white canes, guide dogs, or certified orientation techniques for street crossings and physical safety.
• Facial identification safeguards: In accordance with OpenAI safety guidelines and privacy policies, Be My AI restricts direct identity recognition of unfamiliar individuals to protect personal privacy, describing general visual appearance instead.
• Device hardware compatibility: Mobile applications run on modern iOS (iOS 15 or higher) and Android (Android 8.0 or higher) smartphones. Advanced real-time video features require functional camera and microphone hardware permissions.

Pricing breakdown

Be My Eyes maintains a social impact business model that guarantees free access for individuals while generating revenue through corporate accessibility services.

*Pricing and plan details verified as of September 2026. Personal mobile app access remains permanently free for all blind and low-vision users.

Implementation and onboarding pipeline

Getting started with Be My Eyes takes only a few minutes through a streamlined mobile onboarding sequence.

1. Mobile client download: Download the Be My Eyes application from Apple App Store or Google Play Store. The application automatically configures interface sizing based on device accessibility preferences.
2. Role selection and account setup: Open the application and choose your primary role: I need visual assistance or I would like to volunteer. Complete quick account creation using an email address, Apple ID, or Google login.
3. Camera and microphone permissions: Grant necessary system permissions for camera access, audio microphone input, and notification alerts. Ensure screen reader tools (VoiceOver or TalkBack) are active if speech feedback is required.
4. Executing first visual query or test call: Tap the Be My AI button to capture a test photo of an everyday object, reviewing the detailed spoken description. Alternatively, press the Call Volunteer button to test video and audio connection fidelity with the global network.

Structured editorial verdict

Best for: Blind, low-vision, and visually impaired individuals seeking daily visual independence, alongside enterprises committed to accessible customer service delivery.

Not recommended for: Users seeking a high-speed text generation chatbot, or individuals requiring certified GPS outdoor navigation and physical obstacle collision avoidance.

Learning curve: Extremely low. The user interface features high-contrast buttons, large text elements, and flawless integration with VoiceOver and TalkBack screen readers.

Value threshold and ROI: Incomparable. By providing unlimited multimodal vision AI and human volunteer assistance completely free of charge to visually impaired individuals, Be My Eyes offers life-changing daily utility with zero financial barrier.

Bottom line: Be My Eyes sets the gold standard in assistive technology, combining state-of-the-art GPT-4 vision capabilities with an expansive, compassionate global human volunteer network.