← All companies

Applications / Company profile

Midjourney

What Midjourney does

Midjourney is a self-funded, consumer-focused generative media platform that produces images (and increasingly video) from text and image inputs. The service is accessed primarily through Discord and via a web interface, with paid subscription plans that govern generation capacity (including “Relax Mode”), optional private/“Stealth” access, and higher-tier access to video generation. Midjourney positions itself as an independent research lab with a small distributed team and an actively moderated community, using frequent model iterations and product updates to improve image quality, prompt understanding, and workflow features like references, personalization, and faster generation modes. Recent releases include a new default model cadence (V8.1) and additional model iteration (V8.2), plus features such as Omni-Reference (for preserving referenced entities across scenes), web UX improvements for “alpha” workflows, and expansion beyond images with its V1 video model.

News

Company record

Aug 21, 2026 · Official · MidjourneyChangelog - 8/20/26

Midjourney published a changelog entry describing new alpha-stage UX improvements, including organizing work into folders and sidebar navigation changes.

Jul 24, 2026 · Official · MidjourneyVersion 8.2

Midjourney launched V8.2, describing improvements focused on aesthetics/image quality and personalization behavior.

Jul 23, 2026 · Official · MidjourneyMidjourney’s First Acquisition (Co–Star)

Midjourney announced its first acquisition: Co–Star, stating that Co–Star remains under Banu’s control/direction while joining Midjourney for expanded product and design resources.

Jun 11, 2026 · Official · MidjourneyV8.1 is now the default model on Midjourney!

Midjourney updated the default model from V7 to V8.1 and described changes including faster rendering times, better coherence/prompt adherence, improved text rendering, and HD mode behavior.

Feb 26, 2026 · Official · MidjourneyPersonalization and Web Updates

Midjourney announced a new personalization interface on web, added personalization/moodboards to niji 7, and announced removal (“sunsetting”) of Rooms on web while pointing users toward Discord for events and live-chat.

Feb 26, 2026 · Official · MidjourneyAn update on Rooms

Midjourney removed the Rooms feature from the website, explaining it as an experiment that didn’t scale and stating it is making space to build future experiences for live support and events.

Jan 09, 2026 · Official · MidjourneyNiji V7!

Midjourney announced Niji V7, describing it as its anime-focused model and highlighting improvements in coherence, prompt understanding, text rendering, and performance for sref.

Jun 18, 2025 · Official · MidjourneyIntroducing Our V1 Video Model

Midjourney introduced its V1 video model and described a roadmap idea spanning image models, video models, and building blocks toward more real-time/open-world simulation concepts.

Show 4 earlier updates
Jun 17, 2025 · Official · MidjourneyV7 is now the default model!

Midjourney updated the default model to V7 after testing, describing improvements such as higher image quality, better prompt understanding, Omni-reference and draft mode behavior, and improvements to personalization.

May 01, 2025 · Official · MidjourneyOmni-Reference --oref

Midjourney began testing Omni-Reference, explaining how to use it on web (drag-and-drop into the prompt bar’s omni-reference area) and on Discord (e.g., --oref url and --ow).

Apr 30, 2025 · Official · MidjourneyV7 update, editor, and --exp

Midjourney announced V7-related workflow/product updates, including an editor update and changes involving --exp.

Aug 23, 2024 · Official · MidjourneyThe web is open to all!

Midjourney announced opening up its web platform to everyone, including instructions related to signing in with Discord to merge web image history.

Source map · 3 recurring channels · 17 references

Still resolving: Newsroom

Funding

Private company

Round-level funding appears only when a dated source is attached; the profile does not turn an undisclosed amount into zero.

What it builds

Midjourney Image Service

The core Midjourney image-generation service, available to subscribers through the Midjourney web and Discord experiences.

source ↗
Midjourney Web platform

A custom-built web experience for creating and exploring images, with account sign-in options and web-specific features and workflows.

source ↗
Discord integration (Midjourney bot/server)

Discord-based access where users generate images and interact with the community; the service also supports signing in on web via Discord and using features through Discord commands.

source ↗
Subscription plans (Basic, Standard, Pro, Mega)

Auto-renewing subscription tiers that define generation capacity and feature access, including “Relax Mode” and higher-tier access to video generation and Stealth mode.

source ↗
Stealth Mode (Pro/Mega)

A private mode for keeping generated assets private, available only on Pro and Mega plans.

source ↗
Omni-Reference (V7-era reference system)

A reference system that lets users include an image reference in prompts (via web drag-and-drop or Discord parameters) to preserve referenced characters, objects, vehicles, and non-human creatures across generated scenes.

source ↗
Video generation: V1 Video Model

A video-generation capability introduced as a “V1 video model,” enabling users to animate generated images to produce short video outputs.

source ↗
Model line: V8.1 / V8.2 default model updates

Model updates that Midjourney rolls forward as the default model (including V8.1 and subsequent V8.2 refinements), with changes focused on coherence, prompt adherence, text rendering, speed, and personalization/aesthetics.

source ↗

Milestones & partnerships