What Memories.ai does

Memories.ai is a San Francisco-based AI company building what it describes as the world's first Large Visual Memory Model (LVMM), a foundation architecture that gives AI systems persistent, searchable recall of visual experiences across unlimited timeframes. The LVMM maps video frames, audio features, and text semantics into a unified semantic space; its successor, LVMM 2.0, is optimized for on-device edge deployment via Qualcomm processors. The company's product portfolio spans four layers: (1) core model and APIs, including the LVMM family, a public API platform at api-platform.memories.ai, and the LUCI Desktop on-device research initiative; (2) a suite of 20+ consumer and developer tools covering video search, summarization, transcription, editing, captioning, and platform-specific analyzers for YouTube, TikTok, and Instagram; (3) vertical enterprise solutions for Security and Surveillance, Media and Production, Video Marketing, Sports Analytics, and Robotics; and (4) Research Fellowship and bounty programs to recruit talent.

The company monetizes through a hybrid revenue model: a freemium self-serve platform (100 free credits per month), a Plus subscription at $20 per month for 5,000 credits with rollover, usage-based credit purchases at $0.0046 per credit, custom enterprise contracts with SLA, on-premise and private cloud deployment options, and API and OEM licensing revenue from hardware partners including Qualcomm, NVIDIA, Microsoft (Windows ML), and Ring. Distribution combines a self-serve web app, an API platform, enterprise sales motion for large deployments, and OEM pre-installation on partner hardware.

Founded in 2024 by former Meta Reality Labs researchers Shawn Shen and Ben Zhou, Memories.ai operates as Memories.ai Platforms, Inc., a Delaware-incorporated private company. The company has raised $16 million in total through seed rounds led by Susa Ventures with participation from Samsung Next, Crane Venture Partners, Fusion Fund, Seedcamp, and Creator Ventures. It claims 10,000+ team users, 20+ video platform integrations, and over one million hours of indexed video data. Notable partnerships announced include Samsung (strategic investor and customer), Qualcomm (November 2025), Ring (March 2026, 100M+ camera reach), NVIDIA (March 2026), and Microsoft (June 2026). The company has published four research papers on arXiv and was named to Sifted's 100 Most Promising AI Startups for 2025 and Database Trends and Applications' 30 Startups to Watch in 2026.

Name

Memories.ai

Legal name

Memories.ai Platforms, Inc.

Website

https://memories.ai

Company type

Private

Founded year

2024

Operating status

Operating

Headcount range

11–50 employees

Short description

Memories.ai is a San Francisco-based AI company building the Large Visual Memory Model (LVMM), a foundation architecture that gives AI systems persistent, searchable recall of video across unlimited timeframes, serving enterprise verticals including security, media production, video marketing, sports, and robotics via a hybrid self-serve, API, and OEM licensing model.

Ownership category

akta.pro rank

Product category

AI Video Analytics

NAICS

Software Publishers (5132), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (518)

SIC

Services-Computer Programming, Data Processing, Etc. (7370), Services-Motion Picture & Video Tape Production (7812)

akta.pro primary industry

Real-Time Vision at the Edge (on-device CV/video) (HDAAAJAE)

akta.pro secondary industries

AI/Analytics Video Surveillance (Detection, LPR, People Counting) (HSAHACAF), Digital Asset Management (DAM) for Media/Creative (MPAHAGAB)

Keywords

• AI video analytics
• Visual memory platform
• Video understanding AI
• Multimodal AI
• Video intelligence software

Headquarters

HQ city

San Francisco

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B and B2C

Offering type

Software

Cost components

Technology or R&D, Personnel, Marketing or Sales, Operations, Infrastructure

Revenue model

1. Subscription - Plus Tier: Monthly subscription at $20/month providing 5,000 credits per month with rollover capability. Credits can be used for video analysis, transcription, and other AI features.
2. Credit Purchases: Usage-based credit purchases at $0.0046 per credit. Available in packs of 2,000, 4,000, 10,000, 20,000, and 40,000 credits. One-time purchases for additional consumption beyond subscription limits.
3. Free Tier: Freemium model providing 100 credits per month to attract users and demonstrate value before conversion to paid tiers.
4. Enterprise Custom Contracts: Custom pricing for enterprise deployments including video scriptor agent, custom deployment options (cloud, private cloud, on-premise), SLA guarantees, dedicated technical support, and model fine-tuning.
5. API Access & Licensing: API Platform access for developers and enterprises to embed video intelligence into their own products and workflows. Revenue from developer API usage and technology licensing to hardware partners.

Pricing tiers

Go-to-market motion5 records

Distribution channels7 records

Marketing channels12 records

Core offering

Memories.ai builds and operates a video intelligence platform powered by its proprietary Large Visual Memory Model (LVMM), which converts video data into persistent, searchable, structured visual memory. The company sells self-serve consumer/developer access via a web app and API, subscription tiers for power users, enterprise contracts for large organizations with custom deployment, and OEM technology licensing to hardware partners including Qualcomm, Nvidia, Microsoft, and Ring.

Product overview

Memories.ai is a video intelligence platform built around the world's first Large Visual Memory Model (LVMM), now in its 2.0 iteration. The core LVMM technology converts video data into searchable, structured visual memory, enabling machines to recall visual experiences across unlimited timeframes. The platform operates as a unified system with a Multimodal Data Lake architecture that maps video frames, audio, and text into a shared semantic space. The product portfolio includes: (1) Core LVMM/LVMM 2.0 - the foundational Large Visual Memory Model for persistent, searchable video memory; (2) Project LUCI - an on-device visual AI initiative with LUCI Desktop for PC and wearable integration; (3) Enterprise Solutions - vertical platforms for Security & Safety, Media & Production, Video Marketing, Sports Analytics, and Robotics & AI; (4) Consumer Tools - a suite of 20+ AI-powered video tools including Video Search AI, AI Video Understanding, AI Video Summarizer, Video Editing AI, Video Caption AI, Speaker Recognition, Meeting Transcription, and platform-specific analyzers for YouTube, TikTok, Instagram, and more; (5) Video Marketer and Creator Insight - marketing intelligence and influencer analysis tools; (6) API Platform - developer access to video analysis and visual memory capabilities; and (7) Research Fellowship - academic research program in visual AI. The platform processes video from 20+ platforms natively and supports on-premise, cloud, and edge deployment options.

Differentiator

Problem solved

Functional benefit

Brands

• Project LUCI: A research prototype AI wearable development platform powered by Large Visual Memory Model (LVMM 2.0), designed to enable AI devices to capture, index, remember, and retrieve users' visual experiences and daily moments.
• LVMM

Products and services

• Large Visual Memory Model (LVMM) The foundational Large Visual Memory Model that converts video data into persistent, searchable, structured visual memory by mapping video frames, audio features, and text semantics into a unified semantic space. Powers all Memories.ai products and enterprise solutions.
• LVMM 2.0 On-device visual memory model optimized for edge deployment on Qualcomm processors, enabling sub-second video search with reduced latency, lower cloud costs, and enhanced data privacy by keeping visual data local on devices.
• Project LUCI Local-first visual AI platform that solves 'digital amnesia' by automatically indexing and correlating everything a user sees across their ecosystem using a proprietary 3B-parameter embedding model for 100% on-device processing with zero cloud latency and total privacy.
• Security & Safety Solutions

Quantifiable outcome

• 95% token reduction with maintained reasoning accuracy using MARC compression
• +6 more outcomes

Named customers6 records

Segments6 records

Ideal customer profiles6 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration39 records

AI capability14 records

Feature7 records

Partnerships

Four partnerships are on record, tiered flagship and major.

• MicrosoftflagshipTechnology or Integration · 1 June 2026Project LUCI featured at Microsoft Build 2026. LUCI Desktop uses Windows ML on Qualcomm X2 Elite. Joint demonstration showcasing unified memory across PC, wearables, and IoT.
• RingmajorTechnology or Integration · 17 March 2026Integration bringing LVMM visual memory to over 100 million Ring cameras. Memories.ai available in Ring Appstore. Enables persistent visual memory for the largest camera network globally.
• NvidiaflagshipTechnology or Integration · 16 March 2026Collaboration with Nvidia at GTC conference to develop visual memory technology for AI wearables and robotics. Uses Nvidia's Cosmos Reason 2 model and Metropolis application framework. LVMM optimized for Nvidia AI infrastructure.
• QualcommflagshipTechnology or Integration · 6 November 2025Strategic partnership to run LVMM 2.0 natively on Qualcomm processors starting in 2026. Enables sub-second video search with reduced latency and cloud costs. Targets phones, computers, and wearables with on-device visual AI.

Scale indicators8 records

Recent moves6 records

Expansion highlights7 records

Emerging players

• Roboflow: Roboflow provides a developer platform for building and deploying computer vision models with a focus on workflow and dataset management. Adjacent peer in vision AI tooling, with less emphasis on persistent visual memory but overlapping developer and enterprise buyer personas.
• Twelve Labs competitor: Pixop: Pixop provides AI-powered video enhancement and analysis for media and broadcast customers. Comparable in the Media & Production vertical where Memories.ai offers asset management, search, and AI re-editing capabilities.

Direct peers

• Clarifai: Clarifai offers a computer vision AI platform with model training, inference, and visual search capabilities. Direct peer in the AI vision platform category, serving developers and enterprises that need to build video and image understanding into their products.
• Spot AI: Spot AI provides AI-powered video surveillance and analytics for businesses, transforming camera footage into searchable intelligence. Direct peer in the security and safety vertical where Memories.ai competes on similar AI video search and anomaly detection capabilities.
• Twelve Labs: Twelve Labs builds foundation models for video understanding, offering semantic video search and multimodal analysis APIs. Directly comparable to Memories.ai as both are focused on persistent, searchable video memory for AI systems and serve similar developer and enterprise customers.
• Voxel AI: Voxel applies computer vision to workplace safety and security using existing camera infrastructure. Direct peer competing on similar real-time threat detection and behavioral analysis use cases that Memories.ai targets in its Security & Safety solutions.
• Hive: Hive provides AI-powered visual intelligence APIs for content moderation, object detection, and visual search. Comparable peer offering developer-facing vision AI APIs similar to Memories.ai's API Platform, targeting similar use cases in content analysis and understanding.

Broad incumbents

• Google DeepMind: Google's Gemini family of multimodal models powers video understanding, transcription, and visual analysis at massive scale across Google Cloud and consumer products. Broad incumbent that Memories.ai positions its LVMM performance against and that has substantially larger R&D resources.
• Verkada: Verkada is a leading cloud-managed physical security platform combining cameras with AI-powered analytics. Broader incumbent serving similar enterprise security customers with integrated hardware plus software, where Memories.ai is a software-only layer that integrates with Verkada-adjacent VMS platforms.
• OpenAI: OpenAI builds general-purpose multimodal AI models (GPT-4V, Sora, ChatGPT) that increasingly cover video understanding, transcription, and search. Broad incumbent whose models Memories.ai explicitly benchmarks against — the largest competitive threat to Memories.ai's differentiation.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks6 records

Key highlights7 records

Customer concentration

• Xx.com/memories_ai
• YouTubeyoutube.com/@memories_dot_ai
• LinkedInlinkedin.com/company/memoriesai
• Instagraminstagram.com/memories_dot_ai
• TikToktiktok.com/@memories_dot_ai

Compliance4 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles3 records

Funding overview

Funding rounds1 record

Investors6 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Memories.ai