Midjourney

Midjourney

Overview

Midjourney is a San Francisco-based independent AI research lab founded by David Holz (CEO, also co-founder of Leap Motion) in 2021. Distinguished by its unique product position — image generation distributed primarily through Discord and now its own web app, with a paid subscription model and no free tier — Midjourney has built one of the most profitable and culturally influential generative AI products without external venture funding. The company has consistently set the aesthetic frontier in AI image generation through V1 → V8.1, with each version associated with distinctive visual styles that propagate through online creative communities.

Midjourney V7 (default since June 2025) introduced Draft Mode (10x faster, half-cost) and Omni Reference, with V8 Alpha following in preview. V8.1, released April 30, 2026, is currently described as Midjourney's fastest model. Through 2025–2026, Midjourney has expanded beyond still images into AI video generation, putting the company in direct competition with Runway, Luma, and Kling at the consumer/creative end of the video market.

Key Details

  • Founded: 2021
  • Founder / CEO: David Holz
  • Headquarters: San Francisco, CA
  • Distribution Model: Discord + Midjourney.com web app; paid subscription only (no free tier)
  • Pricing: Plans start at $10/month
  • Strategic Positioning: Aesthetic-frontier image generation; community-driven distribution; expanding into video
  • Website: https://www.midjourney.com

Current Models

  • V8.1 — Released April 30, 2026; Midjourney's fastest model
  • V8 Alpha — Live in preview as of mid-2026
  • V7 — Default since June 17, 2025; introduced Draft Mode (10x faster, half-cost) and Omni Reference
  • V6.1 / V6 / earlier — Prior generations remain available
  • Video Generation — Midjourney has expanded into video generation through 2025–2026, competing with Runway, Luma, Pika, and Kling

Key People

  • David Holz — Founder and CEO; previously co-founded Leap Motion; described in industry coverage as one of the most distinctive product leaders in generative AI

Recent Developments

  • V8.1 Release (April 30, 2026): Midjourney's fastest model to date; live on midjourney.com.
  • V8 Alpha: In preview alongside the V7 default; signaling continued aggressive iteration on the underlying model.
  • V7 Default Promotion (June 17, 2025): V7 introduced Draft Mode (10x faster, half-cost) and Omni Reference for stronger character / object consistency across generations.
  • Video Expansion: Midjourney's expansion into video generation positions the company as one of the few labs operating across both image and video at the consumer/creative tier — a different positioning from FLUX (image-only) and Runway/Luma/Kling (video-first).
  • Pricing Stability: $10/month entry tier preserved through 2025–2026 despite quality leaps — unusual in AI products where pricing typically scales with capability.

Why They Matter

Midjourney is the most distinctive product success story in generative AI: a small independent lab that has consistently set the aesthetic frontier in image generation, built a deeply loyal subscriber community, generated reportedly hundreds of millions in annual revenue, and operated with no external venture funding — all without ever offering a free tier. Where Stable Diffusion / FLUX dominate the open-weight ecosystem and OpenAI's GPT-image / Google's Nano Banana 2 dominate the integrated multi-modal frontier, Midjourney owns the "I want the best aesthetic results, I'm willing to pay, I want a community" position with very little credible competition. The expansion into video generation positions the company to potentially repeat the image-generation playbook in the much-larger video market — though competing with Runway, Luma, Kling, and Veo simultaneously is a substantially harder fight than the image generation race that Midjourney effectively won. Strategically, Midjourney is the proof that independent generative-AI product companies can be commercially durable at small team size and without VC funding, which has implications for how the broader generative-media market evolves.

Last Updated

May 7, 2026