快速入门
This step by step guide covers every Midjourney feature:
- 入门 — Create your discord account and get started
- How to Use Powerful AI Image Editor — Edit generated images directly without leaving Midjourney
- How to Use Model Personalization — Teach the AI your preferred style using rated image pairs
- How to Use AI Art Workflow — Organize your full ai art creation process from prompt to export
- How to Use Upscale and Print — Upscale any image to print-ready resolution in one click
- How to Use AI Logo Design — Generate professional logo concepts with simple text prompts
- How to Use Image Prompting — Use a reference image as the starting point for new creations
- How to Use AI Parameters — Fine-tune every detail with aspect ratio, style, and chaos parameters
- How to Use Dual Image Effects — Blend two images into a single creative output
- How to Use AI Art Progress — Track how midjourney model versions improve your art over time
- 如何使用换脸功能 — Swap faces in generated images while keeping the scene intact
所需时间： 每部影片 5 分钟
本指南还包含以下内容： 专业提示 | 常见错误 | 故障排除 | 定价 | 替代方案
为什么信任本指南
I’ve used Midjourney for over a year and tested every feature covered here. This tutorial comes from real hands-on experience — not marketing fluff or vendor screenshots.
Midjourney is one of the most powerful ai art tools to generate images today.
但大多数用户仅仅触及了它功能的冰山一角。
本指南将向您展示如何使用所有主要功能。
一步一步教你，附带截图和专业技巧。
Midjourney Tutorial
This complete How to Use Midjourney tutorial walks you through every feature, from setup to advanced tips that make you a power user.
旅程中途
Generate stunning ai art from simple text prompts. Midjourney turns words into images in about one minute. Start with the Basic Plan — no design skills needed.
Getting Started with Midjourney
使用任何功能之前，请先完成此一次性设置。
大约需要3分钟。
第一步：创建您的帐户
Go to midjourney.com and click “Sign In” to create your discord account.
You can also join the Midjourney Discord server directly from Discord.
Sign up with your email and confirm your account.
✓ 检查点： 检查你的 收件箱 请发送确认邮件。
Step 2: Choose Your Midjourney Subscription
Pick a paid plan that fits your needs — the basic plan starts at $8 per month.
Midjourney no longer offers a consistent free trial, so a subscription is required.
这就是仪表盘的样子：
✓ 检查点： You should see the main dashboard with the create tab visible.
Step 3: Generate Your First Image
打开 想象 bar, type a text prompt, and press enter to start generating images.
Your first image appears in about one minute as a grid of four options.
✅ 完成： 您已准备好使用以下任何功能。
How to Use Midjourney Powerful AI Image Editor
Powerful AI Image Editor lets you edit generated images directly inside Midjourney without switching tools.
以下是使用步骤。
Step 1: Open Your Generated Image
Click a selected image from your gallery on midjourney.com.
The editor panel opens on the right side.
Step 2: Use the Edit Tools
Select a region, erase an object, or repaint part of the image.
这就是它的样子：
✓ 检查点： You should see the edited area regenerated in real time.
Step 3: Save Your Final Edit
Click the download icon to save the full size version to your device.
✅ 结果： You edited a Midjourney image without leaving the platform.
💡 专业提示： Use the erase tool on small areas first — smaller edits give the Midjourney bot more context for cleaner results.
How to Use Midjourney Model Personalization
Model Personalization lets you teach Midjourney your preferred style so every prompt matches your taste.
以下是使用步骤。
Step 1: Open Your Personalization Settings
Go to midjourney.com and click Personalization in Settings.
You will see a ranking feed of image pairs.
Step 2: Rate Image Pairs
Pick the image you prefer from each pair — this becomes training 数据 for your profile.
这就是它的样子：
✓ 检查点： A progress bar shows how many ratings you completed.
Step 3: Apply Personalization to Prompts
Add the –p flag to any text prompt to activate your personalized style.
✅ 结果： Your Midjourney images now reflect your unique aesthetic.
💡 专业提示： Rate at least 200 pairs for strong results. Remix Mode lets you modify the original prompt while keeping personalized style references intact.
How to Use Midjourney AI Art Workflow
AI Art Workflow lets you organize your ai art creation process from prompt to final export.
以下是使用步骤。
Step 1: Start a New Project
Open the create tab on midjourney.com and enter your text prompt.
Descriptive sentences are better understood by the Midjourney bot than keyword dumps.
Step 2: Generate and Refine
Press enter and wait about one minute for your new grid of images to appear.
这就是它的样子：
✓ 检查点： You see two rows of midjourney images in the output panel.
Step 3: Export Your Best Work
Right click the image you want, save it full size, or share with other users.
✅ 结果： You completed a full workflow from prompt to download.
💡 专业提示： Use folders to sort projects by client or theme — this keeps your ai art organized as your library grows.
How to Use Midjourney Upscale and Print
Upscale and Print lets you upscale any generated image to print-ready resolution.
以下是使用步骤。
Step 1: Select an Image to Upscale
Click U buttons below a generated grid to upscale your selected image.
Each button matches one of the four images in the grid.
Step 2: Choose Upscale Level
Pick standard or high-definition upscale depending on your print size needs.
这就是它的样子：
✓ 检查点： The upscaled image loads at higher resolution in the viewer.
Step 3: Download for Print
Click the download icon or right click and save to get the full size file.
✅ 结果： Your image is print-ready at high resolution.
💡 专业提示： For large prints, use an external upscaler after Midjourney’s built-in upscale to push resolution further without losing details.
How to Use Midjourney AI Logo Design
AI Logo Design lets you generate professional logo concepts using simple text prompts.
以下是使用步骤。
Step 1: Write a Logo Prompt
Describe the logo you want — include the object, style, and color palette.
For example, type ‘minimal fox logo, flat design, orange and white.’
Step 2: Review the Logo Grid
Midjourney returns four logo options in a new grid within one minute.
这就是它的样子：
✓ 检查点： You should see multiple images with distinct logo variations on screen.
Step 3: Refine with Variations
Click V buttons to create variations of your favorite logo with slight variations.
✅ 结果： You have polished logo concepts ready for client review.
💡 专业提示： Add ‘–no text’ to your prompt to remove auto-generated lettering — this gives you a clean icon for custom typography.
How to Use Midjourney Image Prompting
Image Prompting lets you use a reference image as the starting point for new creations.
以下是使用步骤。
Step 1: Upload Your Reference Image
Paste the image URL directly into the prompt field or drag and drop a file.
Midjourney allows image prompts by reading the visual content of your file.
Step 2: Add a Text Description
After the image url, type a text prompt describing changes you want.
这就是它的样子：
✓ 检查点： The imagine bar shows both the image prompt and your text together.
Step 3: Adjust Image Weight
Use the –iw parameter to control how much the reference image influences output.
✅ 结果： Midjourney blended your reference image with your text description.
💡 专业提示： Set –iw to 2 for strong image influence or 0.5 to let the text prompt dominate. Use style reference images for consistent looks across multiple images.
How to Use Midjourney AI Parameters
AI Parameters lets you fine-tune every detail of your image generation with multiple parameters.
以下是使用步骤。
Step 1: Learn Core Parameters
Start with –ar for aspect ratio, –s for stylize value, and –c for chaos.
Adjusting the stylize value influences how closely the Midjourney bot follows text prompts.
Step 2: Add Parameters to Your Prompt
Type your prompt, then append parameters at the end, for example ‘–ar 16:9 –s 250.’
这就是它的样子：
✓ 检查点： The output should reflect your chosen aspect ratio and style intensity.
Step 3: Experiment with Advanced Flags
Try –seed to reproduce results, –raw to strip the default aesthetic filter, or –no to exclude an object.
✅ 结果： You customized your image output with precision using parameters.
💡 专业提示： Keep a cheat sheet of your favorite parameter combos. The –seed parameter lets you generate similar images across sessions.
How to Use Midjourney Dual Image Effects
Dual Image Effects lets you blend two reference images into a single creative output.
以下是使用步骤。
Step 1: Choose Two Source Images
Pick two images that share a theme or contrast in an interesting way.
Upload both image URLs into the prompt field.
Step 2: Run the Blend Command
Use the /blend command or paste both image URLs followed by your prompt.
这就是它的样子：
✓ 检查点： Midjourney shows a grid combining visual elements from both sources.
Step 3: Refine the Blend
Add parameters like –iw to shift focus toward one image or create variations using V buttons.
✅ 结果： You merged two images into a unique blended creation.
💡 专业提示： Pick source images with similar lighting — this helps Midjourney blend them naturally.
How to Use Midjourney AI Art Progress
AI Art Progress lets you track how midjourney model versions improve your art over time.
以下是使用步骤。
Step 1: View Your Generation History
Open your profile on midjourney.com and 浏览 your past generated images.
Every image you generated is saved with its original prompt.
Step 2: Compare Across Model Versions
Re-run an old prompt on the latest model to see how generative ai models evolved.
这就是它的样子：
✓ 检查点： You should see visible quality improvements between older and newer versions.
Step 3: Save Your Best Results
Bookmark or download standout images to track your creative growth.
✅ 结果： You can track how your skills and Midjourney’s models have improved together.
💡 专业提示： Re-run your very first image prompt every few months — comparing outputs shows how fast generative ai is advancing.
How to Use Midjourney Face Swap
换脸 lets you replace faces in a generated image while keeping the rest intact.
以下是使用步骤。
Step 1: Generate a Base Image
Create your starting image using a detailed text prompt with a clear subject.
Make sure the face is clearly visible and well-lit.
Step 2: Upload a Face Reference
Use the image prompt feature to provide a reference face via image url.
这就是它的样子：
✓ 检查点： The prompt field should show the base image reference plus the face source.
Step 3: Generate the Swapped Result
Press enter and review the output — Midjourney merges the face into the scene.
✅ 结果： You swapped the face while preserving the scene and style.
💡 专业提示： Use a high-resolution, front-facing face photo for the cleanest swap. Even lighting produces the most natural results.
Midjourney Pro Tips and Shortcuts
After testing Midjourney for over a year and watching dozens of tutorial videos, here are my best tips.
键盘快捷键
| 行动 | 捷径 |
|---|---|
| Open imagine bar | /想象 |
| Blend two images | /blend |
| View generation settings | /settings |
| Send prompt in direct messages | /imagine in DM |
大多数人错过的隐藏功能
- Remix Mode: Enable it in /settings to modify your original prompt when creating variations — more control over each iteration.
- Prompt Engineering with Weights: Use syntax like ‘sunset::2 ocean::1’ to focus on specific elements in your composition.
- Multi-Prompt Strategy: Separate concepts with commas so the Midjourney bot interprets each element independently.
Midjourney Common Mistakes to Avoid
Mistake #1: Using Vague Single Word Prompts
❌ 错误： Typing one generic word like “dog” and expecting a specific output.
✅ 右图： Use descriptive sentences — specificity improves image generation quality.
Mistake #2: Ignoring Parameters Entirely
❌ 错误： Never adding –ar, –s, or –no to your midjourney prompts.
✅ 右图： Add multiple parameters to control aspect ratio, style, and composition.
Mistake #3: Skipping Upscale Before Downloading
❌ 错误： Saving the small grid preview instead of upscaling first.
✅ 右图： Always upscale your selected image before downloading for highest resolution.
Midjourney Troubleshooting
Problem: Images Not Generating
原因： Your midjourney subscription may have run out of GPU time, called tokens.
使固定： Check your account page for remaining fast hours and upgrade or wait for reset.
Problem: Prompt Blocked or Filtered
原因： Certain words trigger Midjourney’s content filter.
使固定： Rephrase your text prompt using different words and focus on descriptive language.
Problem: Discord Bot Not Responding
原因： The Midjourney server may be under heavy load.
使固定： Refresh Discord, try a different channel, or use midjourney.com directly.
📌 笔记： If none of these fix your issue, contact Midjourney support.
什么是中途之旅？
旅程中途 is a generative ai tool that turns text prompts into stunning images.
Think of it like having an 艺术家 who can paint anything you describe in words.
观看这段快速概览：
它包含以下主要特点：
- Powerful AI Image Editor: Edit generated images directly on the platform.
- Model Personalization: Rate images to teach the AI your preferred style.
- AI Art Workflow: Organize your creation process from basic prompts to export.
- 升级和打印： Upscale any selected image to print-ready quality.
- AI Logo Design: Generate logo concepts using descriptive text prompts.
- 图片提示： Use an image prompt or style reference to guide new output.
- AI Parameters: Control details with parameters like aspect ratio and stylize.
- Dual Image Effects: Blend two images into one unique output.
- AI Art Progress: Track how midjourney model versions improve over time.
- 换脸： Replace faces while keeping the scene intact.
如需完整评测，请参阅我们的 Midjourney review.
旅程中期定价
Here’s what Midjourney costs in 2026:
| 计划 | 价格 | 最适合 |
|---|---|---|
| 基本计划 | $8 | Hobbyists trying AI art for the first time |
| 标准计划 | $24 | Regular creators who need more GPU time |
| 专业计划 | $48 | Power users running multiple images daily |
| 巨型计划 | $96 | Teams and professionals at full scale |
免费试用： Midjourney no longer offers a consistent free trial. A paid plan is required.
退款保证： No official refund policy — review plan details before subscribing.
💰 性价比最高： Standard Plan — enough GPU time for daily image generation without the mega plan price.
中途之旅与替代方案
How does Midjourney compare? Here’s the competitive landscape:
| 工具 | 最适合 | 价格 | 等级 |
|---|---|---|---|
| 旅程中途 | Artistic AI image generation | 每月 8 美元 | ⭐ 4.6 |
| DALL-E 3 | Photorealism | 每月 20 美元 | ⭐ 4.5 |
| 稳定扩散 | 开源 | 自由的 | ⭐ 4.4 |
| Adobe Firefly | Creative Cloud | 每月10美元 | ⭐ 4.3 |
| 莱昂纳多人工智能 | Free-tier speed | 免费/每月 12 美元 | ⭐ 4.4 |
| Canva | Quick design | 免费/每月 13 美元 | ⭐ 4.5 |
快速精选：
- 综合最佳： Midjourney — unmatched artistic quality and active model updates.
- 最佳预算： 蜡笔 — completely free with no account required.
- 最适合初学者： Canva — familiar design interface with built-in 人工智能工具.
- 最适合照片写实风格： DALL-E 3 — sharper realistic images and text rendering.
🎯 中途替代方案
正在寻找 Midjourney 的替代方案？以下是一些最佳选择：
- 🚀 DALL-E 3： OpenAI’s image 发电机 excels at photorealism and text rendering inside images.
- 💰 稳定扩散： Open-source and free to run locally with full control over generative ai models.
- 🎨 Adobe Firefly： Built into Adobe’s creative suite for designers already using Photoshop.
- ⚡ 莱昂纳多人工智能： Real-time image generation with a generous free tier and fast model training.
- 🔒 Canva： Combines AI image generation with a full design platform for beginners.
- 🧠 NightCafe 创始人： Supports multiple AI art algorithms in one platform for experimenting with styles.
- 👶 WOMBO 的《Dream》： 移动优先 AI图像生成器 with one-tap creation for casual users.
- 🏢 Artbreeder: Focuses on image blending and genetic-style art evolution through community tools.
- 🔧 蜡笔： Formerly DALL-E Mini, now fully free with no paid subscription required.
- 🌟 深度梦境生成器： Transforms photos into artistic styles using neural networks and dream-like effects.
- ⭐ StarryAI： Mobile AI art app with daily free credits for testing image generation.
- 🎯 生成的照片： Specializes in AI-generated headshots and stock photos for 商业 使用。
- 💼 DeepAI： Free API-based image generation for developers building custom tools.
- 📊 碧玉 艺术： Jasper’s built-in image tool tied to their AI writing platform.
- 🔥 Photosonic： Writesonic’s AI image tool focused on marketing visuals and blog headers.
- 🚀 艾雅白： Turns selfies into classical art portraits using AI style transfer.
- 💰 照片: Photo editing platform with built-in AI image generation features.
- 🎨 Picsart: Popular mobile editor with AI image generation and thousands of templates.
- ⚡ 德兹戈: Free web-based generator using Stable Diffusion — no sign-up needed.
- 🔒 菲德拉: AI photo enhancement focused on portrait photography and batch processing.
⚔️ 中途之旅对比
以下是 Midjourney 与各竞争对手的对比：
- Midjourney 对阵 DALL-E 3： DALL E is better for photorealism; Midjourney wins on artistic style and composition.
- 中途扩散 vs 稳定扩散： Stable Diffusion is free but needs technical setup. Midjourney produces polished results 即刻.
- Midjourney 与 Adobe Firefly： Adobe Firefly integrates with Creative Cloud. Midjourney offers more artistic range standalone.
- Midjourney 对阵 Leonardo AI： Leonardo AI has a free tier and faster generation. Midjourney delivers more polished output.
- Midjourney 与 Canva 的对比： Canva is better for quick templates. Midjourney is stronger for custom image generation.
- Midjourney vs NightCafe Creator: NightCafe supports many AI models. Midjourney focuses on one engine with deeper quality.
- WOMBO 的 Midjourney vs Dream： Dream by WOMBO is simpler and mobile-first. Midjourney offers far more control.
- Midjourney 对阵 Artbreeder： Artbreeder blends through community tools. Midjourney provides more precise prompt-based creation.
- Midjourney 对阵 Craiyon： Craiyon is free and instant. Midjourney costs more but delivers much higher quality.
- Midjourney 与 Deep Dream Generator： Deep Dream Generator is niche. Midjourney covers a wider range of styles.
- Midjourney 对阵 StarryAI： StarryAI gives free daily credits. Midjourney offers better quality for serious work.
- 旅途中的照片与生成的照片： Generated Photos focuses on headshots. Midjourney handles broader ai art creation.
- Midjourney 对比 DeepAI： DeepAI is API-first for developers. Midjourney is better for visual-interface users.
- Midjourney vs Jasper Art： Jasper Art bundles with writing tools. Midjourney produces higher quality images standalone.
- Midjourney 对比 Photosonic： Photosonic is tied to Writesonic. Midjourney outperforms in quality and community support.
- Midjourney vs AI Gahaku: AI Gahaku is limited to portraits. Midjourney generates any type of image from prompts.
- Midjourney 对阵 Fotor： Fotor combines editing with basic AI. Midjourney provides dedicated, higher-quality generation.
- Midjourney vs Picsart： Picsart adds AI to mobile editing. Midjourney offers deeper creative control overall.
- Midjourney 对阵 Dezgo： Dezgo is free with no account needed. Midjourney delivers significantly better quality.
- Midjourney 对阵 Phedra： Phedra focuses on portrait enhancement. Midjourney handles full-spectrum image generation.
Start Using Midjourney Now
You learned how to use every major Midjourney feature:
- ✅ Powerful AI Image Editor
- ✅ Model Personalization
- ✅ AI Art Workflow
- ✅ Upscale and Print
- ✅ AI Logo Design
- ✅ Image Prompting
- ✅ AI Parameters
- ✅ Dual Image Effects
- ✅ AI Art Progress
- ✅ 换脸
下一步： 选择一项功能，立即试用。
Most people start with Powerful AI Image Editor.
只需不到5分钟。
常见问题解答
Here are the most common Midjourney FAQs answered from hands-on testing.
How does Midjourney personalization work?
Midjourney personalization uses your past ratings as training data to tune its generative AI models to your style. Rate images you like and the Midjourney bot interprets future prompts based on those preferences.
Does Midjourney own my images?
No. With a paid plan you own images you create. Free-tier creations are licensed under Creative Commons. Check the full list of terms on Midjourney’s website for details.
Can I make my Midjourney profile private?
Yes. Go to midjourney.com, open Settings, and toggle Stealth Mode. This hides your midjourney images from other users. Stealth Mode requires a Pro Plan or Mega Plan subscription.
Midjourney 比 DALL-E 更好吗？
Midjourney excels at artistic style and composition. DALL E is stronger at photorealism and text rendering. Your choice depends on whether you focus on ai art or realistic image generation.
Is Midjourney easy to use?
Yes. Type a text prompt, press enter, and your first image appears in about one minute. Basic prompts need only a single word or short sentence to start generating images.
English 
Spanish 
Chinese 
German 
French 
Arabic 
Russian 
Portuguese 
Japanese 
Korean 
Italian 
Turkish 
Dutch 
Polish 
Indonesian 
Thai 
Vietnamese 