TL;DR
- Revenue Milestone: Midjourney’s annual revenue significantly surpassed $200 million in 2023 and has continued growing, CEO David Holz confirmed to The Information.
- Profitable Outlier: Midjourney reached profitability within a year of founding without outside investment, making it one of the leanest AI companies in the industry.
- Hardware Ambitions: The company is exploring hardware projects that may require venture capital for the first time, marking a potential shift from its self-funded identity.
- Legal Exposure: Disney and Universal filed a 110-page copyright infringement lawsuit against Midjourney in June 2025, seeking up to $150,000 per infringed work.
- Competitive Context: Despite operating at roughly 1/80th of OpenAI’s revenue scale, Midjourney is profitable while OpenAI is not expected to break even until 2030.
Midjourney revealed that its annual revenue “significantly surpassed” $200 million in 2023 and has continued growing, CEO David Holz reported in an interview with The Information published March 29, 2026. Profitability makes the AI image generator a rare outlier in a sector where competitors burn through billions chasing scale.
Holz’s disclosure paints a picture of a company operating at a level few expected from a bootstrapped outfit with no outside investors. Meanwhile, Midjourney’s ambitions now extend beyond software into hardware, a capital-intensive move that could force the startup to seek venture funding for the first time, potentially reshaping the lean operating model that made it successful.
The Bootstrapped Anomaly
Founded in 2021 by serial entrepreneur David Holz as an independent research lab, Midjourney reached profitability within a year without taking a dollar of outside investment. That trajectory stands in sharp contrast to competitors like Stability AI, which faced financial struggles and leadership upheaval even as it raised hundreds of millions.
In contrast, Stability AI’s founder Emad Mostaque stepped back in late 2024 amid an $80 million lifeline and mounting AWS debts, while the company itself entered talks for a potential sale earlier that year.
Midjourney was generating that level of ARR with just 40 employees in 2023, a figure that translates to roughly $5 million per head. Per Time’s profile of Holz, revenue pushed to $300 million by 2024, suggesting the company may now generate more than $7 million per employee if headcount has remained lean.
Moreover, Midjourney has amassed over 20 million users on Discord’s largest server, building its audience through a community-driven model rather than enterprise sales teams.
Midjourney’s platform operates on a public-by-default creation model, where images generated by one user can be remixed and built upon by others. That openness has fueled viral growth and community engagement, but it also raises questions about intellectual property rights for paying subscribers who want exclusive control over their work.
Beyond static images, Midjourney continues expanding its product line. Earlier in 2025, it launched its V7 image generation model, maintaining its reputation for photorealistic output and distinctive artistic styles.
Building on this momentum, in July it launched Midjourney TV, a 24/7 stream of videos made with its V1 video model, signaling ambitions well beyond still image generation. In August 2025, Meta licensed Midjourney’s technology for image and video generation, providing enterprise validation and a new revenue stream.
Capital efficiency at this scale positions Midjourney as one of the leanest AI operations in the industry. Meta’s licensing deal demonstrates that even tech giants with their own research labs see value in paying for Midjourney’s output rather than replicating it in-house.
Hardware Ambitions and the Capital Question
Midjourney is now exploring hardware projects that could require the kind of capital its software business has not previously needed. According to The Information, the company may need venture capital to move further into hardware, a shift that would mark a fundamental change in its identity as a self-funded operation.
Hardware interest at Midjourney predates this disclosure. WinBuzzer reported in August 2024 that the company had hired a former Apple Vision Pro engineer to lead a new hardware division, one of several strategic hires that signaled the company’s intent to move beyond pure software.
Furthermore, Holz’s latest comments indicate those plans have progressed to the point where outside funding is under serious consideration. Building physical products would require supply chain management, manufacturing partnerships, and regulatory compliance, all areas where Midjourney has no track record.
For a company that has prided itself on independence, taking venture capital would introduce new dynamics: board seats, growth expectations, and potential pressure to prioritize scale over the deliberate, quality-focused approach that has defined Midjourney’s culture.
Separately, the acquisition question has drawn external attention. Investor Michael Burry, known for his “Big Short” trade against subprime mortgages, publicly suggested Adobe acquire Midjourney to protect its creative software franchises.
“Adobe should buy Midjourney. And any other small creative software company with a creative founder. You have the cash flow to protect your franchises.”
Michael Burry, Founder of Scion Asset Management
Adobe’s in-house AI efforts with Firefly have not sparked strong investor enthusiasm, lending weight to Burry’s argument that acquisition could be a faster path to competitive AI capabilities. Firefly offers a free tier while Midjourney requires a paid subscription starting at $10 per month, yet Midjourney’s perceived quality advantage continues to justify its pricing.
However, whether Midjourney would entertain such an offer while still profitable and fully independent remains unclear, but the rising cost of hardware development could quickly change that calculus.
Competitive Pressures and Legal Battles
The capital question comes against a backdrop of intensifying market competition. According to Data Insights Market, the global AI image generator market was valued at $468 million in 2025, projected to maintain an 11.5% annual growth rate through 2033.
Moreover, DALL-E, Stable Diffusion, and Adobe Firefly all vie for market share, with each platform offering different pricing models, creative capabilities, and integration strategies. Meta’s decision to license Midjourney’s technology rather than build its own underscores the value gap between established generators and newer entrants.
Legal challenges also loom large. In June 2025, Disney and Universal sued Midjourney alleging copyright infringement, filing a 110-page complaint in Los Angeles federal court seeking up to $150,000 per infringed work. Midjourney denies wrongdoing and argues a fair use defense, but the scope of the complaint and the financial resources of the plaintiffs make this a formidable legal challenge.
Holz’s own prior comments on training data could complicate that defense. “It’s just a big scrape of the internet. We weren’t picky,” he previously acknowledged, a statement that plaintiffs could use to undermine fair use arguments.
“If the courts agree, this could set a precedent that fundamentally changes how AI companies train their models.”
Dustin Taylor, IP Attorney (via Forbes report on AI copyright lawsuit against Midjourney)
As a result, a ruling against Midjourney could force the company and its competitors to license training data or rebuild models from scratch, adding material costs to an industry already struggling with profitability. For Midjourney specifically, licensing fees for copyrighted training material could erode the margins that make its bootstrapped model viable, creating yet another pressure point toward outside capital.
The broader AI copyright environment remains unsettled. Multiple lawsuits targeting different AI image generators are working through federal courts simultaneously, and no definitive ruling has emerged on whether training AI models on copyrighted works constitutes fair use. How these cases resolve will shape the cost structure of every company in the space.
OpenAI has become one of the fastest-growing software companies in the technology sector, with enterprise customers driving solid demand. Its annualized revenue reached $25 billion by the end of February 2026, up 17% from $21.4 billion at the end of 2025.
Yet OpenAI is not expected to reach profitability until 2030, spending heavily on infrastructure, talent, and research. Midjourney’s ability to stay profitable at a fraction of that scale, with a fraction of the staff, remains its primary competitive advantage. OpenAI’s revenue dwarfs Midjourney’s by a factor of roughly 80, but profitability inverts that comparison: Midjourney has it, and OpenAI does not.
Holz now faces a three-front challenge: maintaining profitability while funding hardware ambitions, defending against high-profile copyright litigation, and keeping pace with deep-pocketed competitors. How he navigates that convergence of pressures will determine whether Midjourney’s bootstrapped success story becomes a lasting model or a brief chapter in AI’s rapid evolution.
