Hao0321/ai-media-generator

- [`cliprise/awesome-image-to-video-prompts`](https://github.com/cliprise/awesome-image-to-video-prompts) — i2v core principles + 5-part formula + source image checklist

- [`underwoodxie/promptsref-gpt-image-prompts`](https://github.com/underwoodxie/promptsref-gpt-image-prompts) — GPT-Image2 prompt patterns

- [`cliprise/awesome-ai-product-photography-prompts`](https://github.com/cliprise/awesome-ai-product-photography-prompts) — product photography patterns

Researched and consolidated from [YouMind-OpenLab/awesome-seedance-2-prompts](https://github.com/YouMind-OpenLab/awesome-seedance-2-prompts) (CC BY 4.0) — 6 hand-picked Featured + 101 community-validated prompts, last updated 2026-05-18. Al

# Codex, per-user git clone https://github.com/Hao0321/ai-media-generator.git ~/.codex/skills/ai-media-generator

git clone https://github.com/Hao0321/ai-media-generator.git ~/.codex/skills/ai-media-generator

# Claude Code, per-user git clone https://github.com/Hao0321/ai-media-generator.git ~/.claude/skills/ai-media-generator

git clone https://github.com/Hao0321/ai-media-generator.git ~/.claude/skills/ai-media-generator

# Claude Code, project-level git clone https://github.com/Hao0321/ai-media-generator.git ./.claude/skills/ai-media-generator

git clone https://github.com/Hao0321/ai-media-generator.git ./.claude/skills/ai-media-generator

**AI 短劇邊界：** 使用者要做 AI 短劇／漫劇、重生／系統／逆襲／隱藏大佬、分集連載或整季架構時，先由 [`ai-short-drama`](https://github.com/Hao0321/ai-short-drama) 建題材、Bible、pilot、分集狀態與 production pack；本 Skill 只接下游角色圖、場景圖、逐鏡影片／聲音 prompt 與模型執行，不自行重寫季弧。

| 使用者說… | 讀這個檔 | 為什麼 | |---|---|---| | 「AI短劇/AI漫劇/爽劇/重生/系統/逆襲/隱藏大佬/分集連載」 | [`ai-short-drama`](https://github.com/Hao0321/ai-short-drama) | 先鎖故事、角色、揭露梯與 state；媒體 prompt 再回本 Skill |

|---|---|---| | 「AI短劇/AI漫劇/爽劇/重生/系統/逆襲/隱藏大佬/分集連載」 | [`ai-short-drama`](https://github.com/Hao0321/ai-short-drama) | 先鎖故事、角色、揭露梯與 state；媒體 prompt 再回本 Skill | | 「結果有瑕疵/爛掉/不夠精緻/閃爍/變形/塑膠感/文字怪」 | [quality-control.md](references/quality-control.m

| 「AI短劇/AI漫劇/爽劇/重生/系統/逆襲/隱藏大佬/分集連載」 | [`ai-short-drama`](https://github.com/Hao0321/ai-short-drama) | 先鎖故事、角色、揭露梯與 state；媒體 prompt 再回本 Skill |

### OiiOii (https://www.oiioii.ai/) — 2026-04 實測腳本

1. `navigate` `https://www.oiioii.ai/`

### Kling (https://klingai.com/)

**操作步驟：** 1. 導航到 `https://app.klingai.com/global/` 或 `https://klingai.com/`

1. 導航到 `https://app.klingai.com/global/` 或 `https://klingai.com/`

### Suno (https://suno.com/)

**操作步驟：** 1. 導航到 `https://suno.com/create`

1. 導航到 `https://suno.com/create`

### Midjourney (https://www.midjourney.com/)

**操作步驟 (Web UI)：** 1. 導航到 `https://www.midjourney.com/imagine`

1. 導航到 `https://www.midjourney.com/imagine`

### Runway (https://app.runwayml.com/)

中文 UI。 - `https://jimeng.jianying.com/ai-tool/home`

- `https://jimeng.jianying.com/ai-tool/home`

- `https://www.doubao.com/chat/`

### Vidu (https://www.vidu.com/)

**URL:** `https://...`

**URL:** `https://labs.google/fx/tools/flow` (會重導至 `/zh/tools/flow` 或地區別)

**URL:** `https://ideogram.ai/`

**URL:** `https://kling.ai/app/` (新版)，舊 URL `https://app.klingai.com/global/` 會 redirect

從首頁 `https://kling.ai/app/` 任一入口進：

- **中央 main grid** 點 `Video Generation` (座標 ~790, 256) - **或頂部 tab** `Video Generation` (座標 ~237, 22)，**預設進入後這個 tab 就是 active** - URL 變 `https://kling.ai/app/video/new`

- URL 變 `https://kling.ai/app/video/new`

**URL:** `https://www.midjourney.com/imagine` (Web UI) / Discord `#newbies-N` 頻道

| 模式 | URL / 位置 | 優勢 | |---|---|---| | **Web UI** | `https://www.midjourney.com/imagine` | 現代介面、操作直覺 |

|---|---|---| | **Web UI** | `https://www.midjourney.com/imagine` | 現代介面、操作直覺 | | **Discord** | Discord 頻道內 `/imagine prompt: XXX` | 老派、社群感、有些進階功能先在 Discord 上 |

| **Web UI** | `https://www.midjourney.com/imagine` | 現代介面、操作直覺 |

- `--cref` **v7 已失效** — 要用 `--oref` 替代角色一致性 - `--oref` 預設 `--ow 100` 通常不夠，調 **300-600** - **Personalization (`--p`)** 需先到 `https://midjourney.com/rank-v7` 評 ~200 張解鎖

- `--oref` 預設 `--ow 100` 通常不夠，調 **300-600** - **Personalization (`--p`)** 需先到 `https://midjourney.com/rank-v7` 評 ~200 張解鎖 - 多重 `--sref` 加權：`--sref 111 222 333`

- **Personalization (`--p`)** 需先到 `https://midjourney.com/rank-v7` 評 ~200 張解鎖

**URL:** https://www.oiioii.ai/

1. `computer.left_click` 中央 prompt 框 2. `computer.type` 完整 brief (含 `\n` 會自動送出) 3. URL 立刻跳 `https://www.oiioii.ai/space/{uuid}`

3. URL 立刻跳 `https://www.oiioii.ai/space/{uuid}`

每個 brief 送出後會拿到 URL `https://www.oiioii.ai/space/{uuid}`。這個 URL 永久保存所有產出 (劇本、角色圖、場景圖)，使用者登入後隨時回來繼續。demo 中的 space: `e9ae7550-9ea1-42a8-8ee9-072057ee2a02`。

Step 1: tabs_create + navigate https://www.oiioii.ai/  + wait 5 + screenshot

Chrome MCP `javascript_tool` 預設用 `eval()` 不允許 top-level await。**含 `await` 的 JS 必須 wrap：**

// 用 browser_batch 把 navigate 跟 click 合一 [   { name: "navigate", input: { url: "https://www.oiioii.ai/home", tabId: TAB_ID }},

[   { name: "navigate", input: { url: "https://www.oiioii.ai/home", tabId: TAB_ID }},   { name: "javascript_tool", input: {

{ name: "navigate", input: { url: "https://www.oiioii.ai/home", tabId: TAB_ID }},

**URL:** `https://app.runwayml.com/`

**URL:**  - **即夢 AI:** `https://jimeng.jianying.com/ai-tool/home`

- **即夢 AI:** `https://jimeng.jianying.com/ai-tool/home`

- **豆包:** `https://www.doubao.com/chat/`

**URL:** `https://sora.com/` (或 `https://chat.openai.com/sora`) — ⛔ 已關閉

**URL:** `https://suno.com/`

https://suno.com/create  ← Custom Mode

https://suno.com/        ← 對話模式

### Phase 2 — 進 Custom Mode - URL `https://suno.com/create` 直達 (不用繞首頁)

- URL `https://suno.com/create` 直達 (不用繞首頁)

- **v4.5-all ✓ Best free model** — 免費可用   - v4 Pro - 若進首頁 `https://suno.com/` 有 Spring Sale modal (X 關 at ~898, 165)

- 若進首頁 `https://suno.com/` 有 Spring Sale modal (X 關 at ~898, 165)

**URL:** `https://www.vidu.com/` (國際) / `https://www.vidu.studio/` (中國)

- 大氣 / 物理特效全庫 → [vfx-effects.md](vfx-effects.md)｜反瑕疵排查 → [quality-control.md](quality-control.md) - 逐字可貼 prompt → [proven-prompts.md](proven-prompts.md)｜跨模型簽名 → [community-prompt-patterns.md](community-prompt-patterns.md) - [Cinematography 

- 逐字可貼 prompt → [proven-prompts.md](proven-prompts.md)｜跨模型簽名 → [community-prompt-patterns.md](community-prompt-patterns.md) - [Cinematography Lighting Ratios 101 (Wandering DP)](https://wanderingdp.com/cinematography/cinematography-school-l

- [Cinematography Lighting Ratios 101 (Wandering DP)](https://wanderingdp.com/cinematography/cinematography-school-lighting-ratios-101/) · [Rembrandt Lighting Guide (Honcho)](https://thehoncho.app/blog/rembrandt-lighting/) · [Contrast Ratio

- [Subsurface Scattering for Realistic Skin (Hailuo)](https://hailuoai.video/pages/knowledge/subsurface-scattering-realistic-skin-tones-guide) · [Why AI Portraits Look Plastic & How to Fix (Vicsee)](https://vicsee.com/blog/ai-portrait-plast

- [Metricsmule AI Filmmaking Prompts](https://metricsmule.com/ai/ai-filmmaking-prompts/)

- [Director's Guide to AI Video (GitHub)](https://gist.github.com/Dowwie/b1015e1c3b4bffedd1cfc76e45c7da96)

- [Wes Anderson Midjourney Library](https://midlibrary.io/styles/wes-anderson)

- [Hailuo: Film Stock Emulation Guide](https://hailuoai.video/pages/knowledge/film-stock-emulation-ai-video-kodak-fuji-guide)

- [Cinematographers with Recognizable Look](https://www.premiumbeat.com/blog/cinematographers-with-recognizable-look/)

- [Cinematic Lighting Prompts (Atlabs)](https://www.atlabs.ai/blog/improve-your-ai-filmmaking-using-cinematic-lighting-prompts)

- [42 Cinematic AI Composition Prompts](https://www.atlabs.ai/blog/42-cinematic-ai-composition-prompts-guide)

- [No Film School AI Prompts](https://nofilmschool.com/ai-prompts-for-filmmakers)

- [Director Rip-off research](https://nofilmschool.com/which-directors-people-using-ai-rip-off)

- [Luxury Fashion Brands with Cinema Directors](https://luxurysociety.com/en/luxury-fashion-brands-using-cinema-directors-in-video/)

- [Hiro Murai Creative Review](https://www.creativereview.co.uk/hiro-murai-director/)

- [Hiro Murai Directors' Library](https://directorslibrary.com/2013/latest/editorial/interviews-and-talks/director-id-hiro-murai/)

- [Fashion Films Business of Fashion](https://www.businessoffashion.com/videos/technology/fashion-2-0-top-10-fashion-films-of-the-season-5/)

- [TeePee Fashion Directors](https://www.teepeefilms.com/fashion-directors)

- [Lights Film School - Fashion Film](https://www.lightsfilmschool.com/blog/fashion-film-new-music-video)

- [Sora 2 Prompts for AI Video](https://www.youngurbanproject.com/sora-prompts-for-ai-video-generation/)

Source: [`liu-kaining/Awesome-Veo3-Prompts`](https://github.com/liu-kaining/Awesome-Veo3-Prompts) 68⭐ — 31 個 JSON-formatted 範本，雙語版（EN+中文 JSON）。

- **結構：** `[subject], [5-8 details], [lighting], [camera/lens], --ar X:Y --stylize N --style raw --sref/--oref` - **長度：** 30-60 comma chips - **🆕 V7 sweet spot params（2026-05-19 確認 from [`Pixmind-io/awesome-midjourney-v7-example-prompts`](h

- **長度：** 30-60 comma chips - **🆕 V7 sweet spot params（2026-05-19 確認 from [`Pixmind-io/awesome-midjourney-v7-example-prompts`](https://github.com/Pixmind-io/awesome-midjourney-v7-example-prompts)）：**   - **Portrait/Editorial：** `--ar 4:5 --

- **🆕 V7 sweet spot params（2026-05-19 確認 from [`Pixmind-io/awesome-midjourney-v7-example-prompts`](https://github.com/Pixmind-io/awesome-midjourney-v7-example-prompts)）：**

### 🆕 Nano Banana Pro（2026-05-19 補充 from [`Banana-Prompts/awesome-nano-banana-prompts`](https://github.com/Banana-Prompts/awesome-nano-banana-prompts) 51⭐）

> **Source：** [`naqashmunir21/awesome-suno-prompts`](https://github.com/naqashmunir21/awesome-suno-prompts) 37⭐ (CC0)，1000+ genre-organized prompts

**Source attribution：** Distilled from [`xjpp22/awesome--sora-prompts`](https://github.com/xjpp22/awesome--sora-prompts) (open source, 102⭐). Each director entry maps to a **Visual Style Prompt** + **Editing Style Prompt** based on their si

- 完整 source：[`xjpp22/awesome--sora-prompts`](https://github.com/xjpp22/awesome--sora-prompts)（每位導演附參考圖）

- [StudioBinder — Shot Composition Rules](https://www.studiobinder.com/blog/rules-of-shot-composition-in-film/)

- [Awesome AI Video Prompts (GitHub)](https://github.com/geekjourneyx/awesome-ai-video-prompts)

- [Sora 2 Prompting Guide](https://cookbook.openai.com/examples/sora/sora2_prompting_guide)

- [Veo 3 Prompting Guide](https://deepmind.google/models/veo/prompt-guide/)

### ⭐ Seedance 2.0 / 2.5 - **Repo：** [`YouMind-OpenLab/awesome-seedance-2-prompts`](https://github.com/YouMind-OpenLab/awesome-seedance-2-prompts)

- **Repo：** [`YouMind-OpenLab/awesome-seedance-2-prompts`](https://github.com/YouMind-OpenLab/awesome-seedance-2-prompts)

- **內容：** 截至 2026-05-31 repository 顯示 3,552 prompts；數量代表覆蓋量，不等於獨立成功驗證 - **強項：** Bracketed labels 結構、3-shot multi-camera、Audio Profile 4-modal - **Repo 2：** [`AtlasCloudAI/awesome-seedance-2.5-prompts-skills`](https://github.com/AtlasCloudAI

- **強項：** Bracketed labels 結構、3-shot multi-camera、Audio Profile 4-modal - **Repo 2：** [`AtlasCloudAI/awesome-seedance-2.5-prompts-skills`](https://github.com/AtlasCloudAI/awesome-seedance-2.5-prompts-skills)（截至 2026-08-08：150 prompts、33 cat

- **Repo 2：** [`AtlasCloudAI/awesome-seedance-2.5-prompts-skills`](https://github.com/AtlasCloudAI/awesome-seedance-2.5-prompts-skills)（截至 2026-08-08：150 prompts、33 categories、50 previews；provider-aware，2.5 不可用時以 2.0 執行）

### ⭐ Veo 3.1 - **Repo：** [`liu-kaining/Awesome-Veo3-Prompts`](https://github.com/liu-kaining/Awesome-Veo3-Prompts)

- **Repo：** [`liu-kaining/Awesome-Veo3-Prompts`](https://github.com/liu-kaining/Awesome-Veo3-Prompts)

### ⭐ Sora 2（導演 style library） - **Repo：** [`xjpp22/awesome--sora-prompts`](https://github.com/xjpp22/awesome--sora-prompts)

- **Repo：** [`xjpp22/awesome--sora-prompts`](https://github.com/xjpp22/awesome--sora-prompts)

### 🌐 跨平台（Veo / Sora / Runway / Pika / Kling） - **Repo：** [`geekjourneyx/awesome-ai-video-prompts`](https://github.com/geekjourneyx/awesome-ai-video-prompts)

- **Repo：** [`geekjourneyx/awesome-ai-video-prompts`](https://github.com/geekjourneyx/awesome-ai-video-prompts)

### ⭐ Midjourney V7 - **Repo：** [`Pixmind-io/awesome-midjourney-v7-example-prompts`](https://github.com/Pixmind-io/awesome-midjourney-v7-example-prompts)

- **Repo：** [`Pixmind-io/awesome-midjourney-v7-example-prompts`](https://github.com/Pixmind-io/awesome-midjourney-v7-example-prompts)

### Midjourney V5（中文） - **Repo：** [`HebeDich/awesome-midjourney-prompts`](https://github.com/HebeDich/awesome-midjourney-prompts)

- **Repo：** [`HebeDich/awesome-midjourney-prompts`](https://github.com/HebeDich/awesome-midjourney-prompts)

### ⭐ Nano Banana Pro - **Repo 1：** [`Banana-Prompts/awesome-nano-banana-prompts`](https://github.com/Banana-Prompts/awesome-nano-banana-prompts)

- **Repo 1：** [`Banana-Prompts/awesome-nano-banana-prompts`](https://github.com/Banana-Prompts/awesome-nano-banana-prompts)

- **Stars：** 51⭐  **內容：** 14 大類分類（Portrait / Landscape / Architecture / Sci-Fi / Cyberpunk / Fantasy / Animals / Still Life / Food / Fashion / Character / Abstract / Nature / Cityscape）   - **強項：** 短句也能 work + style anchor 列表完整 - **Repo 2：*

- **強項：** 短句也能 work + style anchor 列表完整 - **Repo 2：** [`GarvitOfficial/nanoBananaPrompts`](https://github.com/GarvitOfficial/nanoBananaPrompts)   - **Stars：** 45⭐  **內容：** Seed image + Generated image + Reference image triple comparison（57+

- **Repo 2：** [`GarvitOfficial/nanoBananaPrompts`](https://github.com/GarvitOfficial/nanoBananaPrompts)

### ⭐ Suno v5 - **Repo：** [`naqashmunir21/awesome-suno-prompts`](https://github.com/naqashmunir21/awesome-suno-prompts)

- **Repo：** [`naqashmunir21/awesome-suno-prompts`](https://github.com/naqashmunir21/awesome-suno-prompts)

(cd "$d" && git pull)   else     git clone --depth 1 "https://github.com/$repo" "$d"

else     git clone --depth 1 "https://github.com/$repo" "$d"   fi

git clone --depth 1 "https://github.com/$repo" "$d"

官網 `https://bfl.ai/`。2026 時點主力 **FLUX 1.1 Pro**、**FLUX.1 Kontext (Pro/Max/Dev)**、**Flux Schnell** (快速開源) / **Flux Dev** (開源進階) / **Flux Pro** (API 付費版)。**最強的自然語言 + 寫實 + 文字渲染** 影像模型之一。

- BFL 官網：https://bfl.ai/

- FLUX.1 Kontext 介紹：https://bfl.ai/models/flux-kontext

- Kontext 官方 Prompt Guide (image-to-image)：https://docs.bfl.ml/guides/prompting_guide_kontext_i2i

- Hugging Face Kontext dev：https://huggingface.co/black-forest-labs/FLUX.1-Kontext-dev

- Replicate Kontext Pro：https://replicate.com/black-forest-labs/flux-kontext-pro

- Flux 1 Quickstart (Civitai Edu)：https://education.civitai.com/quickstart-guide-to-flux-1/

- Azure AI Foundry 整合公告：https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/black-forest-labs-flux-1-kontext-pro-and-flux1-1-pro-now-available-in-azure-ai-f/4434659

- Skywork Ultimate Guide：https://skywork.ai/blog/flux-prompting-ultimate-guide-flux1-dev-schnell/

- giz.ai FLUX.1 Prompt Guide：https://www.giz.ai/flux-1-prompt-guide/

- getimg.ai Pro Tips：https://getimg.ai/blog/flux-1-prompt-guide-pro-tips-and-common-mistakes-to-avoid

- **官方部落格（Google Blog）：** https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-omni/

- **Google DeepMind 官方：** https://deepmind.google/models/gemini-omni/

- **Prompt Guide（官方）：** https://deepmind.google/models/gemini-omni/prompt-guide/

- **Flow 操作指南（Google Labs）：** https://labs.google/fx/tools/flow

- **YouTube Shorts Create 整合：** https://www.youtube.com/create

- **對話式編輯最佳實踐（社群）：** https://opus.pro（實驗中文版指南）

官方 https://hailuo.ai/。**當前主力 = Hailuo 2.3 Pro / Std**（MiniMax）。擅長 **微表情 / 角色表演 / 動作戲**，市場認可的「角色演員 AI」，5s 影片畫面內容密度與情緒飽和度頂級。

- 官方：https://hailuo.ai/

- OiiOii 嵌入：https://oiioii.ai/（Hailuo 模型列表）

- 2.3 announcement (MiniMax blog)：https://www.minimaxi.com/（非英文站）

官網 `https://ideogram.ai/`。2026 時點主力 **Ideogram 3.0**（2025-03 發布，全用戶 + iOS App 開放）。**西方模型裡文字渲染最準** 的選擇 — 短文字準確率約 90%、可讀性約 95%（自有 typography 模型），對比 Midjourney 短語約 30%。**英文強、中文弱**。

- 官網：https://ideogram.ai/

- 官方文件總站：https://docs.ideogram.ai/

- Magic Prompt 文件：https://docs.ideogram.ai/using-ideogram/generation-settings/magic-prompt

- Style & Style Reference 文件：https://docs.ideogram.ai/using-ideogram/generation-settings/style-and-style-reference

- Canvas / Magic Fill 文件：https://docs.ideogram.ai/canvas-and-editing/canvas/magic-fill

- Canvas 功能頁：https://ideogram.ai/features/canvas/

- Prompting Guide（官方）：https://docs.ideogram.ai/using-ideogram/prompting-guide/2-prompting-fundamentals

- 定價：https://ideogram.ai/pricing ／ https://docs.ideogram.ai/plans-and-pricing/available-plans

- API：https://ideogram.ai/features/api-pricing

- Ideogram 3.0 發布說明（LearnPrompting）：https://learnprompting.org/blog/ideogram-3-0

- 角色一致性（Scenario 第三方教學）：https://help.scenario.com/en/articles/ideogram-character-single-image-character-consistency/

**Source attribution：** 整合自 [`cliprise/awesome-image-to-video-prompts`](https://github.com/cliprise/awesome-image-to-video-prompts) + [`underwoodxie/promptsref-gpt-image-prompts`](https://github.com/underwoodxie/promptsref-gpt-image-prompts

- [preset-packs.md](../templates/preset-packs.md) — Storyboard / Hero shot 範本 - [external-resources.md](external-resources.md) — Upstream prompt repos catalog - [`cliprise/awesome-image-to-video-prompts`](https://github.com/cliprise/awesome

- [external-resources.md](external-resources.md) — Upstream prompt repos catalog - [`cliprise/awesome-image-to-video-prompts`](https://github.com/cliprise/awesome-image-to-video-prompts) — 原始 i2v 研究參考 - [`underwoodxie/promptsref-gpt-image-p

- [`cliprise/awesome-image-to-video-prompts`](https://github.com/cliprise/awesome-image-to-video-prompts) — 原始 i2v 研究參考

- [`underwoodxie/promptsref-gpt-image-prompts`](https://github.com/underwoodxie/promptsref-gpt-image-prompts) — Oii Image 2 [Best] prompt 範本

官方 `https://kling.ai/app/` (新版，舊 `klingai.com` / `app.klingai.com` 會 redirect) / `https://klingai.kuaishou.com/` (國內版)。**當前主力 = Kling 3.0 / 3.0 Omni (O3) / Motion Control 3.0**（2026-02 發布）。擅長 **精準運鏡**、**人體物理動態**、**區域運動控制 (Motion Brush)**，被公

- 官網（新）：https://kling.ai/app/ ｜ 舊 https://klingai.com/（會 redirect）

- **官方 Motion Control 使用指南：** https://app.klingai.com/global/quickstart/motion-control-user-guide

- **官方 角色一致性指南：** https://app.klingai.com/global/quickstart/ai-video-character-consistency

- 官方 prompt 指南 (ImagineArt 整理)：https://www.imagine.art/blogs/kling-2-1-prompting-guide

- Leonardo.Ai 官方合作指南：https://leonardo.ai/news/kling-ai-prompts/

- Kling 3.0 公式 (glbgpt)：https://www.glbgpt.com/hub/kling-3-0-prompt-guide-for-better-ai-videos

- 2.6 Pro (fal.ai)：https://fal.ai/learn/devs/kling-2-6-pro-prompt-guide

- 運鏡大全 (glbgpt)：https://www.glbgpt.com/hub/kling-ai-camera-movements-explained

- Ambience AI 2026 指南：https://www.ambienceai.com/tutorials/kling-prompting-guide

- [Kuaishou 官方公告](https://ir.kuaishou.com/news-releases/news-release-details/kling-ai-launches-30-model-ushering-era-where-everyone-can-be/)

- [Kling VIDEO 3.0 User Guide](https://app.klingai.com/global/quickstart/klingai-video-3-model-user-guide)

- [Motion Control 3.0](https://kling3.io/motion-control-3-0)

- [Kling 3.0 Omni Guide](https://soravideo.art/blog/kling-3-omni-guide)

- [Kling 2.5 Turbo Prompt Guide (Atlabs)](https://www.atlabs.ai/blog/kling-2-5-turbo-prompting-guide)

- [What is Kling O3 (MindStudio)](https://www.mindstudio.ai/blog/what-is-kling-o3-latest-video-model)

- [Kling 全模型 prompt 指南 (VEED)](https://www.veed.io/learn/kling-ai-prompting-guide)

- [Kling 3.0 Reference / 角色一致性 (MagicHour)](https://magichour.ai/blog/kling-30-reference-guide)

- [Kling AI 定價完整指南 2026 (eesel)](https://www.eesel.ai/blog/kling-ai-pricing)

- [Kling 3.0 多鏡 + 原生音訊完整指南 (Morphic)](https://morphic.com/resources/how-to/kling-3.0)

官網 `https://midjourney.com/`。Discord + Web UI 並行。美感預設值最高的圖像模型。

**三種寫法：** 1. **Code**：`--sref 1234567890`（MJ 內建風格庫，數字碼可收藏/重用） 2. **URL**：`--sref https://...jpg`（自己上傳的圖當風格錨）

1. **Code**：`--sref 1234567890`（MJ 內建風格庫，數字碼可收藏/重用） 2. **URL**：`--sref https://...jpg`（自己上傳的圖當風格錨） 3. **Random**：`--sref random`（隨機抽一個，記下喜歡的 code 收藏）

2. **URL**：`--sref https://...jpg`（自己上傳的圖當風格錨）

prompt words --oref https://.../character.jpg --ow 400 --v 7

到 `https://midjourney.com/rank-v7` 評圖解鎖個人美感 profile（約 5 分鐘 rating；解鎖門檻官方說法在 ~40–200 張間浮動 — 確切數字 **待驗證**）。

Step 3  每個場景同一 oref，只換場景描述：         <角色> in <場景A>, <鏡頭/光>         --oref https://i.mj.run/CHAR.jpg --ow 400 --ar 16:9 --s 200 --v 7

<角色> in <場景A>, <鏡頭/光>         --oref https://i.mj.run/CHAR.jpg --ow 400 --ar 16:9 --s 200 --v 7 Step 4  臉跑掉→--ow 拉到 450–500；風格被洗掉→stylize/ow 同步調

--oref https://i.mj.run/CHAR.jpg --ow 400 --ar 16:9 --s 200 --v 7

the character, standing in a bustling Tokyo night market, neon signs reflecting on wet pavement, handheld candid shot, cinematic --oref https://i.mj.run/my-character.jpg --ow 450 --ar 16:9 --s 200 --v 7

--oref https://i.mj.run/my-character.jpg --ow 450 --ar 16:9 --s 200 --v 7

**官方文件（authoritative）** - Version（各版本現況）：https://docs.midjourney.com/hc/en-us/articles/32199405667853-Version

- Version（各版本現況）：https://docs.midjourney.com/hc/en-us/articles/32199405667853-Version

- Parameter List：https://docs.midjourney.com/hc/en-us/articles/32859204029709-Parameter-List

- Style Reference：https://docs.midjourney.com/hc/en-us/articles/32180011136653-Style-Reference

- Omni Reference：https://docs.midjourney.com/hc/en-us/articles/36285124473997-Omni-Reference

- Character Reference（注意 V7 失效）：https://docs.midjourney.com/hc/en-us/articles/32162917505293-Character-Reference

- Video（V1）：https://docs.midjourney.com/hc/en-us/articles/37460773864589-Video

**官方更新公告（primary）** - V8 Alpha：https://updates.midjourney.com/v8-alpha/

- V8 Alpha：https://updates.midjourney.com/v8-alpha/

- Omni-Reference --oref：https://updates.midjourney.com/omni-reference-oref/

- niji V7：https://updates.midjourney.com/niji-v7/

- V1 Video Model：https://updates.midjourney.com/introducing-our-v1-video-model/

- 更新總頁：https://www.midjourney.com/updates

**社群指南 / 深度評測** - Omni Reference 指南（TitanXT）：https://www.titanxt.io/post/control-your-midjourney-creations-a-guide-to-the-new-omnireference-v7

- Omni Reference 指南（TitanXT）：https://www.titanxt.io/post/control-your-midjourney-creations-a-guide-to-the-new-omnireference-v7

- V8 vs V7 比較（MindStudio）：https://www.mindstudio.ai/blog/midjourney-v8-vs-v7-comparison

- V8 prompting（Blake Crosley）：https://blakecrosley.com/blog/midjourney-v8-prompting

- V8 alpha sref v7 實測（geeky curiosity）：https://geekycuriosity.substack.com/p/midjourney-v8-alpha-the-new-style

- Draft Mode 指南（promptsref）：https://promptsref.com/guide/Exploring-MidJourney-Draft-Mode

- 角色一致性 2026（Medium / Pijush Saha）：https://medium.com/@impijushsaha/how-to-create-consistent-characters-in-midjourney-the-complete-guide-for-2026-405c3bfbb4e1

- niji 7 動畫工作流（DomoAI）：https://domoai.app/blog/niji-7-guide-animate-ai-anime

- Sref Codes Library：https://midjourneysref.com/

- Cheatsheet：https://sref-midjourney.com/cheatsheet

- 2026 Prompt 指南（Printify）：https://printify.com/blog/midjourney-prompts/

| 平台 | URL | 用途 | |---|---|---| | **Gemini App** | `https://gemini.google.com/` | 對話中生圖/編輯 |

|---|---|---| | **Gemini App** | `https://gemini.google.com/` | 對話中生圖/編輯 | | **Google Search AI Mode** | 搜尋列 | 搜尋式觸發 |

| **Gemini App** | `https://gemini.google.com/` | 對話中生圖/編輯 |

| **Google Search AI Mode** | 搜尋列 | 搜尋式觸發 | | **Google Lens** | 手機 app | 圖像分析 + 衍生 | | **Flow** | `https://labs.google/fx/tools/flow` | 切到「圖片」模式 (搭配 Veo 3.1 生影片) |

| **Google Lens** | 手機 app | 圖像分析 + 衍生 | | **Flow** | `https://labs.google/fx/tools/flow` | 切到「圖片」模式 (搭配 Veo 3.1 生影片) | | **Google AI Studio** | `https://aistudio.google.com/` | 開發者測試 |

| **Flow** | `https://labs.google/fx/tools/flow` | 切到「圖片」模式 (搭配 Veo 3.1 生影片) |

| **Google AI Studio** | `https://aistudio.google.com/` | 開發者測試 |

- 官方 DeepMind Prompt Guide：https://deepmind.google/models/gemini-image/prompt-guide/

- Google Cloud Ultimate Prompting Guide：https://cloud.google.com/blog/products/ai-machine-learning/ultimate-prompting-guide-for-nano-banana

- Google Developers Blog 發布公告：https://developers.googleblog.com/en/introducing-gemini-2-5-flash-image/

- Google AI 官方文檔：https://ai.google.dev/gemini-api/docs/models/gemini-2.5-flash-image

- Vertex AI 整合：https://cloud.google.com/blog/products/ai-machine-learning/gemini-2-5-flash-image-on-vertex-ai

- Nano Banana Pro (Gemini 3 Pro Image) 發布：https://blog.google/innovation-and-ai/products/nano-banana-pro/

- Nano Banana Pro Tips (Google Blog)：https://blog.google/products/gemini/prompting-tips-nano-banana-pro/

- Leonardo.Ai Prompt Guide：https://leonardo.ai/news/nano-banana-prompt-guide/

- 75 Prompts + Pro Guide (ImagineArt)：https://www.imagine.art/blogs/nano-banana-pro-prompt-guide

- Travis Nicholson Style List (Medium)：https://travisnicholson.medium.com/complete-list-of-prompts-styles-for-nano-banana-ai-images-ee2229f1c48c

- Nano Banana vs Seedream vs Flux 2 (WaveSpeedAI)：https://wavespeed.ai/blog/posts/seedream-5-0-vs-nano-banana-pro-gpt-image-flux-klein-qwen-image-comparison-2026/

OiiOii (官網 `https://www.oiioii.ai/`) 是一個 **多智能體協作** 的 AI 動畫/影片創作平台。你丟一張圖 (或一個想法)，它內部會派出多個 AI 角色 (藝術總監、編劇、角色設計師、場景師、剪輯師、音效師) 分工完成整部動畫。核心概念：**「Upload just one image → full music MV」**。

1. 登入 `https://www.oiioii.ai/` — 登入後會導向 `/home`

- 官網：https://www.oiioii.ai/

- 深度評測 (Kingy AI)：https://kingy.ai/ai/oiioii-oiioii-ai-a-deep-dive-review-the-ai-animation-studio-that-wants-to-replace-your-entire-production-team/

- Pollo AI 評測 + 免費試用入口：https://pollo.ai/m/oiioii-ai, https://pollo.ai/hub/oiioii-review

- NavTools 簡介：https://navtools.ai/tool/oiioii-ai

官網 `https://runwayml.com/` ・ App `https://app.runwayml.com/`。好萊塢等級工作流，定位是 **AI video studio**（生成 + 編輯 + 表演捕捉一條龍），不是單純的 t2v 玩具。

**官方：** - 官網：https://runwayml.com/

- 官網：https://runwayml.com/

- App：https://app.runwayml.com/

- 定價：https://runwayml.com/pricing

- Changelog：https://runwayml.com/changelog

- Research / Gen-4.5：https://runwayml.com/research/introducing-runway-gen-4.5

- Research / Gen-4：https://runwayml.com/research/introducing-runway-gen-4

- Research 總覽：https://runwayml.com/research

**官方教學（Help Center / Academy）：** - Creating with Gen-4 Video：https://help.runwayml.com/hc/en-us/articles/37327109429011-Creating-with-Gen-4-Video

- Creating with Gen-4 Video：https://help.runwayml.com/hc/en-us/articles/37327109429011-Creating-with-Gen-4-Video

- Creating with Gen-4 Image References：https://help.runwayml.com/hc/en-us/articles/40042718905875-Creating-with-Gen-4-Image-References

- Aleph Prompting Guide：https://help.runwayml.com/hc/en-us/articles/43277392678803-Aleph-Prompting-Guide

- Creating with Aleph：https://help.runwayml.com/hc/en-us/articles/43176400374419-Creating-with-Aleph

- Aleph + Reference Image：https://help.runwayml.com/hc/en-us/articles/44609246167059-Controlling-Aleph-edits-with-a-Reference-Image

- Performance Capture with Act-Two：https://help.runwayml.com/hc/en-us/articles/42311337895827-Performance-Capture-with-Act-Two

- Act-Two 多角色對話：https://help.runwayml.com/hc/en-us/articles/41748090660499-Creating-Multi-Character-Dialogues-with-Act-Two

- Creating with Keyframes：https://help.runwayml.com/hc/en-us/articles/34170748696595-Creating-with-Keyframes

- How do credits work：https://help.runwayml.com/hc/en-us/articles/15124877443219-How-do-credits-work

- Gen-4 section 總覽：https://help.runwayml.com/hc/en-us/sections/39888423025683-Gen-4

- Runway Academy：https://academy.runwayml.com/ways-to-use-runway

- Academy / Aleph 影片轉換：https://academy.runwayml.com/tutorial/how-to-transform-videos

- ByteDance Seedance 2.0 official launch: <https://seed.bytedance.com/blog/seedance-2-0-official-launch>

- BytePlus Dreamina Seedance 2.0 prompt guide: <https://docs.byteplus.com/en/docs/ModelArk/2222480>

- BytePlus video generation API: <https://docs.byteplus.com/en/docs/modelark/1520757>

- BytePlus Seedance 2.5 launch landing: <https://www.byteplus.com/en/contact-us/ai-seedance2-5-official>

- TikTok Symphony + Dreamina Seedance 2.0: <https://ads.us.tiktok.com/help/article/about-tiktok-symphony-and-dreamina-seedance-2-0>

- TikTok Symphony generation and prompt tips: <https://ads.us.tiktok.com/help/article/how-to-use-ingredients-to-video-in-symphony-creative-studio?lang=en>

- AtlasCloudAI, 150 prompts / 50 previews / provenance fields: <https://github.com/AtlasCloudAI/awesome-seedance-2.5-prompts-skills>

- Reddit observational bedroom slice-of-life: <https://www.reddit.com/r/seedance2pro/comments/1vj5vd8/how_to_create_a_candid_bedroom_sliceoflife_videos/>

- Reddit MiniDV coffee vlog: <https://www.reddit.com/r/Seedance_AI/comments/1vhv8g8/built_a_full_minidv_camcorder_coffee_vlog_grinder/>

- Reddit 3D animation physical comedy: <https://www.reddit.com/user/AssociationHead6964/comments/1vj43zv/seedance_25_prompt_3d_animation/>

- Reddit two-reference dialogue comparison: <https://www.reddit.com/r/Seedance_AI/comments/1vi3miq/seedance_2_vs_seedance_25_same_prompt_same/>

- Reddit POV action one-take prompt: <https://www.reddit.com/r/HiggsfieldAI/comments/1vi2zce/i_generated_this_video_using_seedance_25_and/>

- ZeroLu X/WeChat-sourced prompt collection: <https://github.com/ZeroLu/awesome-seedance>

- YouMind 2.0 prompt library: <https://github.com/YouMind-OpenLab/awesome-seedance-2-prompts>

- 抖音：詳細導演 prompt vs 極簡結果 prompt：<https://www.douyin.com/shipin/7431031956181288994>

- Bilibili Seedance 2 提示詞系列：<https://www.bilibili.com/video/BV1PjEd6yEc4/>

- Bilibili 高密度時間碼／運鏡案例：<https://www.bilibili.com/video/BV1pSZWBJEup/>

- 知乎多模態 prompt 方法（搜尋入口，規格需回查官方）：<https://www.zhihu.com/search?q=Seedance%202.0%20%E6%8F%90%E7%A4%BA%E8%AF%8D>

- ByteDance Seedance 2.0 launch: <https://seed.bytedance.com/blog/seedance-2-0-official-launch>

- BytePlus Seedance 2.0 prompt guide（2026-07-31 update）: <https://docs.byteplus.com/en/docs/ModelArk/2222480>

- BytePlus 2.5 launch contact page: <https://www.byteplus.com/en/contact-us/ai-seedance2-5-official>

- Dreamina 2.5 generator（頁面同時標示 coming soon；能力需 gate）: <https://dreamina.capcut.com/seedance/seedance-2-5>

- Dreamina prompt guide: <https://dreamina.capcut.com/seedance/seedance-2-5-prompt>

- Dreamina floor-plan workflow: <https://dreamina.capcut.com/seedance/seedance-2-5-floor-plan-to-video>

- Dreamina cinematic ad examples: <https://dreamina.capcut.com/seedance/seedance-2-5-cinematic-ad-generator>

- Dreamina UGC workflow: <https://dreamina.capcut.com/seedance/seedance-2-5-ai-ugc-ads>

- AtlasCloud 150 prompts／50 previews；明列 2.0 executable default 與 2.5 capability gate: <https://github.com/AtlasCloudAI/awesome-seedance-2.5-prompts-skills>

- YouMind 2.0 prompt library: <https://github.com/YouMind-OpenLab/awesome-seedance-2-prompts>

- ZeroLu X／微信來源庫: <https://github.com/ZeroLu/awesome-seedance>

- Reddit 2.0 vs 2.5 animation comparison: <https://www.reddit.com/r/Seedance_AI/comments/1vi3miq/seedance_2_vs_seedance_25_same_prompt_same/>

- Reddit product-ad workflow discussion: <https://www.reddit.com/r/Seedance_AI/comments/1vf3ot8/making_30second_product_ads_with_seedance_25/>

官方來源：<https://seed.bytedance.com/blog/seedance-2-0-official-launch>

BytePlus 已有 **Seedance 2.5 official launch** 頁，可確認名稱／上線行銷：<https://www.byteplus.com/en/contact-us/ai-seedance2-5-official>。但截至研究日，公開 ModelArk prompt guide／API 詳細限制仍主要列 2.0。**不要把 30 秒、4K、50 references 或局部編輯寫成所有 2.5 入口的共同保證。** 先做 [seedance-2-

**Source：** [`YouMind-OpenLab/awesome-seedance-2-prompts`](https://github.com/YouMind-OpenLab/awesome-seedance-2-prompts)（CC BY 4.0；截至 2026-05-31 repository 顯示 3,552 prompts）。篇數代表覆蓋量，不代表 3,552 次獨立成功驗證；2.5 新案例與去重規則見 [seedance-2-5-community-p

- ByteDance Seedance 2.0 官方發布：https://seed.bytedance.com/blog/seedance-2-0-official-launch

- ByteDance Seedance 2.0 官方頁：https://seed.bytedance.com/seedance2_0

- BytePlus Seedance 2.5 official launch：https://www.byteplus.com/en/contact-us/ai-seedance2-5-official

- BytePlus Seedance 2.0 prompt guide：https://docs.byteplus.com/en/docs/ModelArk/2222480

- 官方 Prompt Guide (BytePlus)：https://docs.byteplus.com/en/docs/ModelArk/1631633

- Seedance 1.0 Lite Guide：https://docs.byteplus.com/en/docs/ModelArk/1587797

- 民間整理 (Atlabs 2026)：https://www.atlabs.ai/blog/ultimate-seedance-1-pro-prompting-guide

- WaveSpeedAI (Seedance 2.0 trailer page)：https://wavespeed.ai/landing/seedance

- Scenario help：https://help.scenario.com/en/articles/seedance-models-the-essentials

- Seedance 官方入口 (第三方彙整)：https://www.seedancepro.net/seedance/prompt-guide

- JoyPix AI Lite 指南：https://www.joypix.ai/tutorials/how-to-write-seedance-1.0-lite-prompt/

### 官方 - Seedream 4.0：https://seed.bytedance.com/en/seedream4_0

- Seedream 4.0：https://seed.bytedance.com/en/seedream4_0

- Seedream 4.5：https://seed.bytedance.com/en/seedream4_5

- Seedream 5.0 Lite：https://seed.bytedance.com/en/seedream5_0_lite

- Seedream 4.0 發布公告：https://seed.bytedance.com/en/blog/seedream-4-0-officially-released-beyond-drawing-into-imagination

- BytePlus / ModelArk 官方 Prompt Guide (4.0–4.5)：https://docs.byteplus.com/en/docs/ModelArk/1829186

### Prompt / 教學指南 - fal.ai Seedream v4.5 Prompt Guide：https://fal.ai/learn/devs/seedream-v4-5-prompt-guide

- fal.ai Seedream v4.5 Prompt Guide：https://fal.ai/learn/devs/seedream-v4-5-prompt-guide

- WaveSpeedAI 4.5 完整指南：https://wavespeed.ai/blog/posts/seedream-4-5-complete-guide-2026/

- WaveSpeedAI 4.0→5.0 Complete Tutorial：https://wavespeed.ai/blog/posts/seedream-4-0-to-5-0-complete-tutorial-image-generation-editing/

- magichour 4.0 Reference/Editing Guide：https://magichour.ai/blog/seedream-edit-guide

- Seedream Best Practices 2026：https://evolink.ai/blog/seedream-prompt-guide-best-practices-2026

- 5.0 Lite Guide (scholarviz)：https://scholarviz.com/blog/seedream-5-0-lite-complete-model-guide

### 價格 - Seedream Pricing Guide 2026 (evolink)：https://evolink.ai/blog/seedream-pricing-guide-2026

- Seedream Pricing Guide 2026 (evolink)：https://evolink.ai/blog/seedream-pricing-guide-2026

- Seedream Pricing Guide (ImagineArt)：https://www.imagine.art/blogs/seedream-pricing-guide

- Price Per Token — Seedream 4.5：https://pricepertoken.com/image/model/bytedance-seedream-4-5

官方 `https://openai.com/sora/`。OpenAI 的文生影片旗艦，2026 時點主力 **Sora 2 / Sora 2 Pro**，首創與 **原生同步音訊** + **社群 App (類 TikTok feed)** 整合，標誌功能是 **Cameo（客串）**。

**官方** - OpenAI Sora 2 發表：https://openai.com/index/sora-2/

- OpenAI Sora 2 發表：https://openai.com/index/sora-2/

- **Sora 停運說明（必讀）**：https://help.openai.com/en/articles/20001152-what-to-know-about-the-sora-discontinuation

- Sora Release Notes：https://help.openai.com/en/articles/12593142-sora-release-notes

- Cameo 使用（generating content with cameos）：https://help.openai.com/en/articles/12435986-generating-content-with-cameos

- 官方 Sora 2 Prompting Guide (cookbook)：https://cookbook.openai.com/examples/sora/sora2_prompting_guide

- 開發者版 cookbook：https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide

- Creating videos with Sora：https://help.openai.com/en/articles/12460853-creating-videos-with-sora

- Generating videos on Sora：https://help.openai.com/en/articles/9957612-generating-videos-on-sora

**第三方 / 教學** - Sora 2 Prompting Guide 中文鏡像：https://soratoai.com/en/docs/guides/sora-2-prompting-guide/

- Sora 2 Prompting Guide 中文鏡像：https://soratoai.com/en/docs/guides/sora-2-prompting-guide/

- fal.ai — How to write prompts for Sora 2：https://fal.ai/learn/devs/how-to-write-prompts-sora-2

- Azure AI Foundry — Sora 2 video generation：https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/video-generation

- Sora 2 vs 2 Pro（MindStudio）：https://www.mindstudio.ai/blog/sora-2-vs-sora-2-pro-upgrade-worth-it

- Tips for 2026 (WaveSpeedAI)：https://wavespeed.ai/blog/posts/sora-2-prompting-tips-better-videos-2026/

- Tutorial (DataCamp)：https://www.datacamp.com/tutorial/sora-ai

- [Veo 3 DeepMind Prompt Guide](https://deepmind.google/models/veo/prompt-guide/)

- [Veo 3.1 Google Cloud Ultimate Guide](https://cloud.google.com/blog/products/ai-machine-learning/ultimate-prompting-guide-for-veo-3-1)

- [Sora 2 OpenAI Cookbook](https://cookbook.openai.com/examples/sora/sora2_prompting_guide)

- [100+ Veo 3 Examples (GVN)](https://geekvibesnation.com/google-veo-3-prompts-100-tested-examples-that-actually-work-2026/)

Stability AI `https://stability.ai/`。開源家族主要版本：

- Stability AI Learning Hub - SD 3.5 Prompt Guide：https://stability.ai/learning-hub/stable-diffusion-3-5-prompt-guide

- getimg.ai Prompt Weights：https://getimg.ai/guides/guide-to-stable-diffusion-prompt-weights

- MimicPC SD 3.5 Guide：https://www.mimicpc.com/learn/stable-diffusion-3-5-prompt-guide

- Stable Diffusion Art (通用)：https://stable-diffusion-art.com/prompt-guide/

- stable-diffusion-web 通用教學：https://stable-diffusion-web.com/prompt-guide

- stabledifffusion.com Full Guide：https://stabledifffusion.com/guides/sd-3-5-prompt-guide

- HyprLab SD 3.5 Docs：https://docs.hyprlab.io/browse-models/model-list/stability/stable-diffusion-3.5

- VEED SD3 Guide：https://www.veed.io/learn/stable-diffusion-3-prompts

- SeaArt SD 3.5：https://docs.seaart.ai/guide-1/6-permanent-events/high-quality-models-recommendation/stable-diffusion-3.5

- Complete Negative Prompts Guide (imagetoprompt)：https://www.imagetoprompt.dev/blog/negative-prompts-stable-diffusion/

官網 `https://suno.com/`。2026 時點主力 **Suno v5.5**（2026-03-26 釋出）。文字描述 + 結構化歌詞 → 完整歌曲（人聲 + 編曲 + 混音 + 母帶）。Suno 的核心差異化：**Style / Lyrics 兩個獨立欄位**、**方括號 meta tags 結構標記**、**Voices（聲線一致性）** 與 **Studio 生成式 DAW**（12 軌 stem 分軌）。

**官方：** - 官網：https://suno.com/

- 官網：https://suno.com/

- Pricing：https://suno.com/pricing

- v5.5 公告（Voices / Custom Models / My Taste）：https://suno.com/blog/v5-5

- v5.5 更新說明（Help）：https://help.suno.com/en/articles/11362305

- Introducing v4.5（blog）：https://suno.com/blog/introducing-v4-5

- Introducing v4（Remaster + ReMi）：https://suno.com/blog/v4

- Remaster（Help）：https://help.suno.com/en/articles/8105281

- Studio 介紹（Help）：https://help.suno.com/en/articles/7940161

- Studio 匯出（Help）：https://help.suno.com/en/articles/8128193

**進階指南（第三方）：** - v5.5 Reference: Meta Tags / Style-of-Music（Blake Crosley）：https://blakecrosley.com/guides/suno

- v5.5 Reference: Meta Tags / Style-of-Music（Blake Crosley）：https://blakecrosley.com/guides/suno

- v5.5 Guide: Voices / Custom Models / My Taste（HookGenius）：https://hookgenius.app/learn/suno-v5-5-guide/

- v5.5 Prompts: what works now（Song AI Farm）：https://www.songaifarm.com/blog/suno-prompts-v5-5

- Suno Studio Tutorial 2026（HookGenius）：https://hookgenius.app/learn/suno-studio-tutorial/

- Suno Studio Guide 2026（Undetectr）：https://undetectr.com/blog/suno-studio-guide

- Stems → DAW（Ableton/Logic/FL，Undetectr）：https://undetectr.com/blog/suno-stems-daw-workflow

- Ultimate v4.5 How-To: Personas/Extend/Cover（Civitai）：https://civitai.com/articles/14849

- Covers Guide 2026（Jack Righteous）：https://jackrighteous.com/en-us/blogs/guides-using-suno-ai-music-creation/suno-ai-covers-guide-v4-transform-your-songs-by-style

- Meta Tags & Song Structure（Jack Righteous）：https://jackrighteous.com/en-us/pages/suno-ai-meta-tags-guide

- 500+ Metatags（OpenMusicPrompt）：https://openmusicprompt.com/blog/suno-ai-metatags-guide

- 300+ Style Tags 分類庫（HookGenius）：https://hookgenius.app/learn/suno-style-tags-guide/

- Free vs Pro vs Premier 2026（HookGenius）：https://hookgenius.app/learn/suno-free-vs-pro-comparison/

- v5 Review（Undetectr）：https://undetectr.com/blog/suno-v5-review

官方 DeepMind `https://deepmind.google/models/veo/`。**市場第一個原生生成同步音訊** (對話、環境音、音效、配樂) 的影片模型。2026 時點主力 **Veo 3.1**。

官方來源：[Introducing Gemini Omni (blog.google)](https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-omni/) · [Gemini Omni (DeepMind)](https://deepmind.google/models/gemini-omni/) · [Omni prompt guide](https://deepmin

- 官方 DeepMind Prompt Guide：https://deepmind.google/models/veo/prompt-guide/

- Google Cloud Ultimate Guide for 3.1：https://cloud.google.com/blog/products/ai-machine-learning/ultimate-prompting-guide-for-veo-3-1

- Vertex AI 官方文件：https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/video-gen-prompt-guide

- Replicate Guide：https://replicate.com/blog/using-and-prompting-veo-3

- LTX Studio 範例：https://ltx.studio/blog/veo-prompt-guide

- 100+ 實測範例 (GVN)：https://geekvibesnation.com/google-veo-3-prompts-100-tested-examples-that-actually-work-2026/

- Beginner Guide 2026：https://www.veo3ai.io/blog/veo-3-beginners-guide-2026

- Veo 3.1 on ImagineArt：https://www.imagine.art/blogs/veo-3-1-prompt-guide

- [Segmind - Professional VFX with AI](https://blog.segmind.com/how-to-create-professional-vfx-with-ai-video-generators/)

- [Human Academy: VFX with AI](https://www.humanacademy.ai/en/blog/vfx-com-ai)

- [Awesome AI Video Prompts (GitHub)](https://github.com/geekjourneyx/awesome-ai-video-prompts)

- [Controlling Next-Gen Video AI (Medium)](https://medium.com/@creativeaininja/how-to-actually-control-next-gen-video-ai-runway-kling-veo-and-sora-prompting-strategies-92ef0055658b)

- [AI VFX with Kling Tutorial](https://www.youtube.com/watch?v=rVxNBgtP_bs)

- [Mootion AI Visual Effects](https://www.mootion.com/use-cases/en/AI-visual-effects-generator)

官網 `https://www.vidu.com/` (國際) / `https://www.vidu.studio/` (中國)。生數科技主打 **Reference-to-Video (多主體參考一致性)** 與 **動漫 / 二次元**，是這兩條賽道目前最強的開放模型之一。

**官方：** - 官網：https://www.vidu.com/

- 官網：https://www.vidu.com/

- Vidu Q3 介紹頁：https://www.vidu.com/vidu-q3

- Q1 動漫官方指南：https://www.vidu.com/blog/vidu-q1-ai-2d-anime-guide

- GPT-4o + Vidu 角色動畫教學：https://www.vidu.com/blog/create-and-animate-ai-character-with-gpt4o-and-vidu

- 定價頁：https://www.vidu.com/pricing

- API 定價：https://platform.vidu.com/docs/pricing

**發布新聞：** - Q2 Ref2V 發布 (PRNewswire)：https://www.prnewswire.com/news-releases/vidu-launches-q2-reference-to-video-pioneering-a-new-era-of-high-consistency-and-creative-control-302590002.html

- Q2 Ref2V 發布 (PRNewswire)：https://www.prnewswire.com/news-releases/vidu-launches-q2-reference-to-video-pioneering-a-new-era-of-high-consistency-and-creative-control-302590002.html

- Q2 全面升級 / 5 分鐘故事 (aibase)：https://news.aibase.com/news/22138

- SCMP 報導 (challenge Sora)：https://www.scmp.com/tech/tech-trends/article/3329800/chinese-ai-start-shengshu-unveils-vidu-q2-challenge-openais-sora

**第三方解析 / 評測：** - Q3 是什麼 + 1/30 發布 (Cutout.pro)：https://www.cutout.pro/learn/blog-what-is-vidu-q3/

- Q3 是什麼 + 1/30 發布 (Cutout.pro)：https://www.cutout.pro/learn/blog-what-is-vidu-q3/

- Q3 深度 (CometAPI)：https://www.cometapi.com/what-is-vidu-q3/

- Q3 pricing & credits (promeai)：https://www.promeai.pro/blog/vidu-q3-pricing-credits/

- Vidu pricing 2026 拆解 (Flowith)：https://flowith.io/blog/vidu-pricing-2026-free-vs-pro-vs-enterprise-cost/

- Q1 動漫評測 (ReelMind)：https://reelmind.ai/blog/vidu-q1-dynamic-anime-multi-reference-video

- Scenario 模型總覽：https://help.scenario.com/en/articles/vidu-models-the-essentials/

- WaveSpeedAI Vidu 集合：https://wavespeed.ai/collections/vidu

- Z.AI Q1 開發文件：https://docs.z.ai/guides/video/vidu-q1

**Source:** [YouMind-OpenLab/awesome-seedance-2-prompts](https://github.com/YouMind-OpenLab/awesome-seedance-2-prompts) (CC BY 4.0, attribution required)

**Use case:** Romance、純愛、表情張力、教室、暗戀情境 **Target:** Seedance 2.0 pro，15s，16:9，720p+ **Author:** AIGC｜阳家豪 ([@JiahaoYang_art](https://x.com/JiahaoYang_art))

**Author:** AIGC｜阳家豪 ([@JiahaoYang_art](https://x.com/JiahaoYang_art))

**Use case:** 高奢時尚、超寫實奇幻、流體渲染、視覺特效 **Target:** Seedance 2.0 pro，15s，16:9 **Author:** John ([@johnAGI168](https://x.com/johnAGI168))

**Author:** John ([@johnAGI168](https://x.com/johnAGI168))

**Use case:** 療癒、生活感、ASMR、Slow living、商業形象短片 **Target:** Seedance 2.0 pro，15s，16:9 **Author:** John ([@johnAGI168](https://x.com/johnAGI168))

**Author:** John ([@johnAGI168](https://x.com/johnAGI168))

**Use case:** 動漫真人化、武打/呼吸法、特效對決、爆炸光效 **Target:** Seedance 2.0 pro，15s，16:9 **Author:** John ([@johnAGI168](https://x.com/johnAGI168))

**Author:** John ([@johnAGI168](https://x.com/johnAGI168))

**Use case:** Music Video、街頭嘻哈、反差人物、Trap beat、quoted dialogue + lip-sync **Target:** Seedance 2.0 pro，15s，16:9 **Author:** 松果先森 ([@songguoxiansen](https://x.com/songguoxiansen))

**Author:** 松果先森 ([@songguoxiansen](https://x.com/songguoxiansen))

**Use case:** 賽車、極速、Fast & Furious 風、霓虹都市、NOS 爆炸瞬間 **Target:** Seedance 2.0 pro，15s，16:9 **Author:** Pierrick Chevallier | IA ([@CharaspowerAI](https://x.com/CharaspowerAI))

**Author:** Pierrick Chevallier | IA ([@CharaspowerAI](https://x.com/CharaspowerAI))

**Use case:** 武術、寫實風格、specific 焦段 token、動作運鏡 **Target:** Seedance 2.0 pro，15s，16:9 **Author:** Jahan Zaib ([@jzaib4269](https://x.com/jzaib4269))

**Author:** Jahan Zaib ([@jzaib4269](https://x.com/jzaib4269))

**Use case:** Sports broadcast、live event、引用 commentator dialogue **Target:** Seedance 2.0 pro，15s，16:9 **Author:** UxUi Tega ([@Tegadesigns](https://x.com/Tegadesigns))

**Author:** UxUi Tega ([@Tegadesigns](https://x.com/Tegadesigns))

**Use case:** 風格化 anime、quoted dialogue 雙語 subtitle、Slice of life **Target:** Seedance 2.0 pro，15s，9:16 vertical **Author:** PromptlyAI ([@PromptlyAI_YT](https://x.com/PromptlyAI_YT))

**Author:** PromptlyAI ([@PromptlyAI_YT](https://x.com/PromptlyAI_YT))

**Use case:** Epic battle、IMAX 質感、9:16 vertical、Format anchor 重度使用 **Target:** Seedance 2.0 pro，15s，9:16 **Author:** Elara Noor ([@ElaraNoorAl8](https://x.com/ElaraNoorAl8))

**Author:** Elara Noor ([@ElaraNoorAl8](https://x.com/ElaraNoorAl8))