Midjourney
AI text-to-image and text-to-video generation platform — the Midjourney models (v6/v6.1, v7, and the recently-shipped video models) produce category-defining stylised image and video output distributed primarily through Discord and the Midjourney web platform. Founded 2021 in San Francisco by David Holz (previously co-founded Leap Motion); famously bootstrapped, profitable and rejected all VC funding; named-press triangulation places revenue at approximately $500M in 2025 (up from approximately $300M in 2024), 21M+ registered Discord users and a lean team of approximately 107–163 employees, with implied revenue-per-employee exceeding $3M.
The Business
Midjourney is an AI text-to-image and text-to-video generation platform — the Midjourney models (v6/v6.1, v7 and the recently-shipped video models through 2025) produce category-defining stylised image and video output distributed primarily through Discord (where Midjourney’s bot anchors the most-populated server in the platform’s history) and through the Midjourney web platform that expanded through 2024–2025. The company was founded in 2021 in San Francisco by David Holz, who previously co-founded Leap Motion (2010–2019) — the gesture-recognition pioneer — and frames Midjourney as an independent research lab aimed at expanding the imaginative powers of the human species. Midjourney is famously bootstrapped: the company has taken zero external venture-capital funding through 2025 and was reportedly profitable within its first year of operation. Named-press triangulation places revenue at approximately $500M in 2025 (up from approximately $300M in 2024), 21M+ registered Discord users with 1.2M–2.5M daily active users as of June 2025, and a lean team of approximately 107–163 employees implying revenue-per-employee exceeding $3M — among the most capital-efficient generative-AI operations globally. The most-recent named-press triangulation places implied Midjourney valuation at approximately $10B as a market-implied figure rather than a priced round.
Customers and Distribution
Midjourney’s primary published commercial signals depend on named-press triangulation rather than company disclosures. The customer mix is dominated by individual consumers, creator-professionals, advertising-and-marketing studios using Midjourney for ideation and production, and game-design and concept-art workflows. Distribution sits across two principal motions: the Discord bot distribution (where Midjourney’s server is the most-populated in the platform’s history and the principal consumer-acquisition funnel) and the Midjourney web platform that expanded through 2024–2025 as the second distribution layer plus the principal monetisation surface. Pricing is on Basic, Standard, Pro and Mega tiers in the $10–$120/month range with consumption-based generation limits per tier. The bootstrapped capital structure plus the 21M+ Discord user base plus the implied $3M+ revenue-per-employee economics form an unusually capital-efficient consumer-AI distribution profile that VC-backed competitors cannot easily replicate. Founder David Holz’s recognition in the TIME 100 Most Influential People in AI 2025 list is the principal recent endorsement on the operational-continuity and founder-credibility lens.
Model Strategy
Midjourney is a Verticals-first play under the IM Framework eight-trajectories taxonomy as it applies to AI image-and-video generation: the strategic bet is that vertical depth on the frontier-image-and-video research stack — the proprietary v6/v6.1, v7 and video models, the stylised-output category positioning, the Discord-anchored consumer distribution, and the bootstrapped capital-efficiency discipline — beats generalist platform-bundled creative-AI from Adobe Firefly, OpenAI DALL-E and Sora, and Google Imagen and Veo on the consumer-creator distribution layer. The foundation-model stack is proprietary on the image-and-video generation side and the company explicitly positions the research-team focus as the defensibility moat against frontier-AI-incumbent substitution from the platform-AI labs that ship bundled-in creative-AI capability. The Plus trajectory secondary is validating the bootstrapped-frontier-AI-research architecture as a category — Midjourney is the worked example of how a focused frontier-research bench can compound revenue per employee above $3M without external capital, against materially larger venture-backed peers in text-to-image and text-to-video. The defensibility profile is anchored on three pillars: the Discord-anchored community as the consumer-distribution moat (the most-populated Discord server in the platform’s history), the bootstrapped capital structure as the unit-economics moat (no VC-led capex burn pressure), and the proprietary frontier-research bench as the model-IP moat (v6/v6.1, v7 and the video models). The text-to-video frontier-research cadence (Sora 2, Veo 3, Runway Gen-4, Pika, FLUX-Video) is the principal structural challenge across 2025–2026.
At A Glance
The Numbers
Annualised revenue
Registered users
Headcount (FTE)
Funding to date
Leadership Team
Midjourney is privately held and famously private about its leadership and operating disclosures. Founder David Holz is the only widely-public leadership figure; CFO, CTO and CRO appointments are not separately disclosed in primary sources. Named-press triangulation places the team at 107–163 employees as of 2025 with the team concentrated on the frontier-research bench plus the platform-and-infrastructure functions. We decline to publish specific named-executive entries beyond David Holz pending fresh primary disclosures and reference his TIME 100 Most Influential People in AI 2025 profile and Contrary Research’s business breakdown as the canonical published-leadership references.
IM Framework Scoring
IM’s structured assessment of Midjourney’s competitive position. The summary below is the headline; expand “Show the full analyst-grade analysis” near the bottom for the per-dimension reasoning and evidence. Methodology →
Funding History
| Date | Round | Raised | Post-money | Lead investor(s) |
|---|---|---|---|---|
| 2021 | Bootstrapped — founder-funded | Founder-funded | — | — (no external VC) |
Midjourney has famously taken zero external venture-capital funding through 2025. The company is bootstrapped and was reportedly profitable within its first year of operation per Contrary Research and Sramana Mitra coverage. Founder David Holz has been consistent across the press cycle (TIME 100 AI 2025, Contrary Research, Sramana Mitra) on the deliberate rejection of VC capital. The most-recent named-press triangulation places implied Midjourney valuation at approximately $10B as a market-implied figure rather than a priced round — we mark this as a market-derived reference rather than a formal valuation event. There is no priced round, no Series A/B/C cycle and no formal cap-table disclosure in primary sources.
Competitive Landscape
| Competitor | Positioning | Distribution edge | Threat profile |
|---|---|---|---|
| Adobe Firefly ((NASDAQ: ADBE)) |
Commercial-safe text-to-image, vector and video generation trained on Adobe Stock and licensed content; positioned as the IP-clean enterprise creative-AI option. | Embedded in Photoshop, Illustrator, Premiere and Express across the Creative Cloud installed base; enterprise licensing through existing Adobe contracts. | High — Adobe Firefly ships native text-to-image generation into the Creative Cloud installed base with deep distribution into design-and-creative procurement; the principal head-to-head on enterprise creative-AI procurement at materially larger distribution scale than Midjourney’s Discord-anchored consumer model. |
| OpenAI (DALL-E, Sora, ChatGPT image-generation) ((OpenAI)) |
Native multimodal image and video generation inside ChatGPT — positioned as a feature of the general-purpose assistant rather than a creative-pro tool. | Inside ChatGPT’s 700M+ weekly active user funnel and the OpenAI API; bundled into Plus, Team, Enterprise and Edu tiers. | High — OpenAI’s image and video generation surfaces (DALL-E inside ChatGPT, Sora text-to-video) ship native creative-AI capability into the largest consumer-AI distribution funnel globally; the structural substitution risk on the consumer image-and-video category. |
| Google (Imagen, Veo) ((NASDAQ: GOOGL)) |
Frontier image (Imagen 3/4) and video (Veo 2/3) models with hyperscale research bench; positioned across Gemini consumer, Workspace and Vertex AI surfaces. | Bundled into Gemini Advanced, Google Workspace AI tiers, Vertex AI for enterprise and the standalone Whisk / Flow / Vids consumer apps; YouTube / Google Photos as latent distribution. | High — Google’s Imagen text-to-image and Veo text-to-video models ship into Workspace, Cloud and consumer AI surfaces with the resource-and-distribution advantages of a platform-AI incumbent. |
| Black Forest Labs (FLUX) | Open-weights frontier image models from the ex-Stable Diffusion research team; explicit research-bench challenge to Midjourney’s proprietary stack, with FLUX.1 Pro / Dev / Schnell tiers. | Open-weights distribution via Hugging Face plus an API for hosted inference; licensed into xAI (Grok image generation) and a growing list of downstream creative-AI products. | Medium-high — FLUX text-to-image models from Black Forest Labs (ex-Stable Diffusion team) compete on the frontier-image-research surface with open-source distribution and an explicit research-bench challenge to Midjourney’s proprietary positioning. |
| Runway / Pika / Stability AI | Runway and Pika lead on AI video (Gen-4 / Pika 2); Stability ships open-source Stable Diffusion across image, video and audio — the diversified frontier-research challenge across modalities Midjourney is now entering. | Runway via direct subscription plus Hollywood-studio partnerships; Pika via consumer-app virality; Stability via open-weights distribution on Hugging Face and a hosted API. | Medium-high — Runway Gen-4 and Pika compete head-to-head on the frontier-video research surface; Stability AI Stable Diffusion competes on the open-source text-to-image distribution layer. The text-to-video competitive cadence is the principal frontier-research challenge for Midjourney across 2025–2026. |
Potential Risks
Frontier-AI-incumbent substitution — Adobe, OpenAI, Google
Adobe Firefly, OpenAI DALL-E / Sora and Google Imagen / Veo ship native creative-AI capability into materially larger consumer and enterprise distribution funnels. The structural substitution risk is that frontier-AI-incumbent bundles win the procurement default on the longer time horizon — particularly on enterprise creative-procurement (Adobe) and consumer-AI-default (OpenAI) segments. The bull case is that Midjourney’s category-defining stylistic output and the Discord-anchored creator community generate durable category preference that platform-bundled creative-AI cannot match; the bear case is that platform-native generative-AI bundles compress the addressable market as text-to-image and text-to-video become commodity capabilities.
Discord-distribution dependency
Midjourney’s distribution has historically been concentrated on Discord, with the web platform increasingly the second distribution layer through 2024–2025. The bull case is that the Discord-anchored community is the structural defensibility moat — the most-populated Discord server in the platform’s history compounds against frontier-incumbent platform-AI; the bear case is that Discord platform-rule cycles, server-architecture changes or distribution-strategy shifts at Discord could compress Midjourney’s distribution at the bot-and-channel layer. The web-platform expansion through 2024–2025 is the explicit de-risking move.
Text-to-video frontier-research cadence
Midjourney shipped initial text-to-video models through 2025; the frontier-video competitive set (OpenAI Sora 2, Google Veo 3, Runway Gen-4, Pika, Black Forest Labs FLUX-Video) competes at materially larger compute and research budgets. The bull case is that Midjourney’s bootstrapped-discipline plus the proprietary research-bench depth produce competitive frontier-video output without venture-backed capex burn; the bear case is that the text-to-video frontier-research cadence will require capex at levels that strain the bootstrapped capital structure.
Structural privacy of operating disclosures
Midjourney famously does not formally disclose revenue, headcount, user counts or strategic metrics in primary sources. Most figures cited across the press cycle (the $500M 2025 revenue, the $300M 2024 revenue, the 21M+ Discord user count, the 107–163 headcount, the $10B implied valuation) depend on named-press triangulation rather than company-disclosed primary sources. The bull case is that operational privacy is a deliberate competitive choice that does not constrain the underlying business performance; the bear case is that the privacy constrains analyst confidence and rules out formal valuation framing pending fresh primary disclosures.
Founder-concentration and bootstrapped-capital-structure dynamics
Midjourney is founder-led with David Holz as the only widely-public leadership figure; no CRO, CFO or CTO is publicly named. The bootstrapped capital structure means the company does not face VC-led pressure but also means there is no diversified institutional governance layer. The bull case is that founder-concentration plus bootstrapped discipline are themselves the strategic advantages relative to VC-backed peers and that David Holz’s TIME 100 AI 2025 recognition validates the operational continuity; the bear case is that single-founder governance carries multi-quarter risk on succession-and-continuity questions that the company does not address in primary sources.
Recent IM Coverage
- Horizontal AI Applications — sector landing May 2026.
- IM Framework Methodology May 2026.
Show recent press coverage of Midjourney
- 2025 — David Holz: The 100 Most Influential People in AI 2025 (TIME)
- 2025 — Report: Midjourney Business Breakdown & Founding Story (Contrary Research)
- Mar 2025 — Ultralight Startup Midjourney Bootstraps to $200M in Revenue (Sramana Mitra)
- 2025 — Midjourney revenue, funding & news — analyst breakdown (Sacra)
- 2024 — David Holz | Biography, Midjourney Founder & CEO (Aituts)
- 2026 — Midjourney — homepage and showcase (Midjourney)
Show the source register for the figures on this page
IM operates a primary-source-where-possible discipline. The figures above come from:
- Revenue: Midjourney does not formally disclose revenue in primary sources. Named-press triangulation places revenue at approximately $500M in 2025 per Contrary Research’s business breakdown and consistent named-press cycle commentary, up from approximately $300M in 2024 per the same cohort of analyst-and-business-press coverage. Sramana Mitra’s bootstrapped-startups column referenced a $200M revenue milestone in early-2025. We anchor the headline at approximately $500M for 2025 with the caveat that all figures depend on named-press triangulation rather than primary company disclosures, and decline to publish a precise exact revenue figure for the current period pending a fresh primary disclosure.
- Users — Discord-and-web: Midjourney disclosed 21M+ registered Discord users as of June 2025 per named-press triangulation, with 1.2M–2.5M daily active users. The web platform is the second distribution layer with growth through 2024–2025. We reference Contrary Research’s business breakdown and the named-press cycle as the canonical disclosed-user references.
- Headcount: Named-press triangulation places Midjourney’s headcount at approximately 107–163 employees as of 2025, growing from 11 employees in 2022. The implied revenue-per-employee figure exceeds $3M — among the most capital-efficient generative-AI operations globally. Midjourney does not separately disclose headcount in primary sources. We reference Contrary Research as the canonical published-headcount reference.
- Funding to date: Midjourney has famously taken zero external venture-capital funding through 2025. The company is bootstrapped and was reportedly profitable within its first year of operation per Contrary Research and Sramana Mitra’s bootstrapped-startups coverage. Founder David Holz has been consistent across the press cycle on the deliberate rejection of VC capital. There is no priced round, no Series A/B/C cycle and no formal cap-table disclosure in primary sources. The most-recent named-press triangulation places implied Midjourney valuation at approximately $10B as a market-implied figure rather than a priced round.
Methodology & Disclaimer
For metric definitions, source-tier hierarchy, and decline-to-publish rules, see the tracker methodology. Confidence dots (• green / • amber / • red) follow the same convention as the AI Tracker.
Spotted a figure you believe is wrong? Send corrections to info@informationmatters.net.
Information Matters Framework scores are the considered opinion of the IM team — human and AI — applied to publicly-available evidence under a disclosed methodology. They are not statements of fact about the companies scored and they are not investment advice.
