DeepMark

What DeepMark does

DeepMark is a private AI company that builds invisible, permanent watermarking technology for digital media. Founded in 2023 and incorporated as DeepMark Inc. in the United States, the company originated from a university research lab in Montenegro founded by PhD researchers Slavko Kovačević and Kosta Pavlović. The platform embeds imperceptible identifiers into audio, image, and video content using deep learning watermarking models combined with classical signal-processing techniques such as spectrogram analysis, peak detection, and combinatorial hashing, alongside neural embeddings built on foundation models including MERT, Mu, and wav2vec 2.0. DeepMark is developing a 128-bit neural audio watermarking algorithm that materially exceeds the 16-20 bit capacity typical of neural systems, and is participating in the C2PA coalition to support content provenance standards.

The product portfolio consists of the core DeepMark watermarking platform and the Vigil Audio Watermarking SDK, a C/C++ library for real-time embedding and detection with cross-language bindings for Python, Rust, and Unity. Target customers include generative AI platforms needing compliance with the EU AI Act and US NO FAKES Act, media production companies, digital publishers, and developer platforms integrating watermarking into their products. DeepMark's research has been cited as pioneering and state-of-the-art by Meta, Microsoft, Sony, and Adobe.

DeepMark operates on an API-first, developer-centric go-to-market motion, offering SDKs and REST API endpoints in Python, Node.js, Java, Go, .NET, Ruby, and PHP, supplemented by enterprise field sales targeting platforms requiring compliance-ready GenAI solutions. Revenue is generated through SDK/API licensing and enterprise contracts on subscription/recurring terms, though pricing is quote-based and not publicly disclosed. The company raised approximately $878,617 from Lighthouse Ventures in September 2025, with additional non-dilutive support from the Montenegro Innovation Fund, Katapult Accelerator, and AWS as a growth partner.

Name

DeepMark

Legal name

DeepMark Inc.

Website

https://deepmark.me

Company type

Private

Founded year

2023

Operating status

Operating

Headcount range

1–10 employees

Short description

DeepMark is an AI-powered watermarking company that embeds invisible, permanent identifiers into audio, image, and video content for provenance and regulatory compliance, serving generative AI platforms, media producers, and developers via SDKs and REST APIs.

Ownership category

akta.pro rank

Product category

Digital Watermarking Software

NAICS

Software Publishers (513210)

SIC

Services-Computer Programming, Data Processing, Etc. (7370)

akta.pro primary industry

Content Authenticity, Watermarking & Deepfake Detection (HDAAAKAK)

akta.pro secondary industries

Content Identification & Matching (Fingerprinting, ID Resolution) (MPAHALAE), Forensic Watermarking (Session/User Watermarking, Payload Design) (MPAHAMAA), Watermarking (Forensic/Visible) & Traceability (MPAHALAL)

Keywords

• AI watermarking technology
• Digital content authentication
• Audio fingerprinting
• C2PA content provenance
• Synthetic media compliance

Headquarters

HQ city

San Francisco

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

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

Revenue model

1. SDK and API Licensing: DeepMark provides developer SDKs and REST API endpoints for watermarking integration. Revenue is generated through licensing of technology to developers and platforms integrating watermarking into their products.
2. Enterprise Solutions: Enterprise field sales motion targeting businesses requiring compliance-ready GenAI solutions, likely involving custom contracts and support agreements.

Go-to-market motion2 records

Distribution channels2 records

Marketing channels4 records

Core offering

DeepMark is an AI-powered digital watermarking platform that embeds invisible, permanent, and unbreakable identifier tags into audio, image, and video content without affecting perceptual quality. The platform provides developer SDKs and a REST API for integrating watermarking and audio fingerprinting into third-party products, along with C2PA soft-binding capabilities, enabling provenance, regulatory compliance (EU AI Act, US NO FAKES Act), and resilient tracking across digital media.

Product overview

DeepMark is a unified AI-powered watermarking platform offering invisible, permanent ID embedding across audio, image, and video content. The product portfolio consists of the core DeepMark watermarking platform complemented by the Vigil Audio Watermarking SDK for developers. DeepMark's technology enables compliance with AI regulations (EU AI Act, US NO FAKES Act), content authenticity verification, copyright protection, and traceability for generative AI platforms, media producers, and digital publishers. The company is expanding into video integrity verification, telco fraud prevention, anti-piracy solutions, and secure communications.

Differentiator

Problem solved

Functional benefit

Brands

• Vigil SDK: C/C++ library for real-time audio watermark embedding and detection using deep learning models.

Products and services

• DeepMark Watermarking Platform AI-powered digital watermarking platform that embeds invisible, permanent, and unbreakable ID tags into audio, image, and video content. Enables authenticity verification, regulatory compliance with the EU AI Act and US NO FAKES Act, content traceability, and copyright protection for AI-generated and digital media assets. Targeted at AI platforms, media producers, and digital publishers.
• Vigil Audio Watermarking SDK C/C++ library for real-time audio watermark embedding and detection using deep learning models. Provides imperceptible digital watermarking with cross-language support via C API bindings for Python, Rust, Unity, and other languages. Features include real-time streaming embedding, offline and real-time detection, and a built-in thread tuner for performance optimization. Targeted at developers integrating watermarking into production products.
• C2PA Soft-Binding Resolution API Open-source C2PA soft-binding solution with manifest ingestion for declaring watermarks within C2PA manifests, plus a soft-binding resolution API for verifying and resolving watermark-based content bindings. Targeted at companies integrating C2PA content authenticity standards with soft-binding technologies.

Quantifiable outcome

• Watermark remains invisible, unalterable, and impossible to remove once embedded
• +2 more outcomes

Named customers4 records

Segments5 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

AI capability7 records

Feature8 records

Partnerships

One partnership is on record.

• C2PA (Coalition for Content Provenance and Authenticity)coreStrategic or Co-development PartnerDeepMark actively participates in C2PA working group meetings and aims to join the official soft binding solutions list. The company is building an open-source C2PA soft-binding solution and contributes to the content authenticity ecosystem.

Scale indicators3 records

Recent moves5 records

Expansion highlights5 records

Broad incumbents

• Google DeepMind (SynthID): Google DeepMind's SynthID embeds imperceptible watermarks directly into AI-generated text, audio, image, and video produced by Google's models. It is the most direct competing platform-level watermarking stack to DeepMark and serves many of the same Generative AI compliance use cases, but is integrated natively into Google services rather than sold as a third-party SDK.
• Microsoft (Azure AI Content Safety / Watermarking): Microsoft ships content provenance and watermarking capabilities as part of Azure AI Content Safety and its Copilot ecosystem, addressing C2PA-style authenticity and EU AI Act compliance for enterprise customers. It overlaps with DeepMark's regulatory-compliance positioning but as part of a much broader enterprise AI safety portfolio.
• Adobe (Content Credentials): Adobe Content Credentials is a C2PA-based provenance and watermarking system applied across Photoshop, Firefly, and Creative Cloud. It competes directly with DeepMark's content authenticity story for digital publishing and creative workflows, but is bundled into Adobe's creative suite rather than offered as a developer SDK.
• OpenAI (text/image watermarking research): OpenAI has publicly researched and deployed provenance signals (C2PA metadata, image provenance classifiers, watermarking in DALL-E outputs). It competes with DeepMark on the GenAI content authenticity narrative but as an in-platform feature inside its own models rather than as a third-party SDK.

Direct peers

• Truepic: Truepic provides C2PA-aligned content authenticity and invisible watermarking APIs for image and video, including its 'Controlled Capture' technology. It is one of the closest direct peers to DeepMark in terms of delivering tamper-evident provenance and watermarking as a developer-facing service for regulated and high-trust use cases.
• Digimarc: Digimarc is an established digital watermarking vendor covering audio, image, and packaging with both visible and forensic watermarks. It overlaps with DeepMark's forensic/audio watermarking offering but is a much larger, publicly traded incumbent with a focus on supply chain and media piracy use cases.
• Audible Magic: Audible Magic provides audio fingerprinting and content identification services for broadcast monitoring, rights management, and UGC platforms. Its content identification and matching capabilities are closely comparable to DeepMark's audio fingerprinting stack, though it does not focus on GenAI provenance.
• Pex: Pex offers audio and video fingerprinting plus an IP rights engine that identifies reused content across platforms, including AI-generated derivative works. It competes with DeepMark's audio fingerprinting and content identification capabilities, particularly in anti-piracy and rights enforcement workflows.
• IMATAG: IMATAG provides invisible watermarking and content tracking APIs for images and video, serving media monitoring, copyright protection, and leakage detection use cases. It is comparable to DeepMark as a developer-focused watermarking platform, though it has historically emphasized tracking rather than GenAI provenance.

Emerging players

• Voiceflow / Resemble AI (synthetic media provenance): Resemble AI builds watermarking and detection tools (Resemble Detect) for AI-generated voice and audio, directly targeting deepfake and synthetic media compliance. It overlaps with DeepMark's AI audio watermarking offering and shares the same enterprise compliance buyer, but focuses almost exclusively on voice.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat4 records

Key risks6 records

Key highlights7 records

Customer concentration

• LinkedInlinkedin.com/company/deepmark-me
• Xx.com/deepmark_me

Revenue estimate

Valuation estimate

Number of profiles

Profiles3 records

Funding overview

Funding rounds1 record

Investors1 record

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 DeepMark