
專項更強提供注重隱私的開發者 API 與開源模型,但比起擴展媒體處理流程,它更優先支援無限制的文字生成。
Scenario: Provides generative production tools and APIs for game assets, but has a narrower focus on 3D and graphic design. Scenario 每月有 336 次造訪,月增率 +37.7%,定價模式為 Free tier plus paid subscriptions (Starter $15/mo, Pro $45/mo, Team $75/mo) and custom Enterprise pricing。WaveSpeedAI 每月有 2,164,832 次造訪,成長率 +22.9%,定價模式為付費。

用你自己的美術素材訓練 AI 模型,在一條管線裡完成圖像、影片、3D 與音訊生產
WaveSpeedAI is an AI media generation platform that exposes 1,000+ image, video, audio, 3D and language models through a single unified API. Developers call any model from Node, Python or cURL, while creators use hosted generators, Cinema Studio, a desktop app and CLI/MCP tooling to produce images, cinematic video, speech and music without managing infrastructure.
| 面向/指標 | ![]() | |
|---|---|---|
| 官方網域 | scenario.gg | wavespeed.ai |
| 定價模式 | Free tier plus paid subscriptions (Starter $15/mo, Pro $45/mo, Team $75/mo) and custom Enterprise pricing | 付費 |
| 每月造訪次數 | 336 (336) | 216.48萬 (2,164,832) |
| 月流量變化量 | +92 (+37.7%) | +40.35萬 (+22.9%) |
| 主要市場與市佔 | 🌐 全球 (—) | 🇺🇸 美國 (17.5%) |
| 每月直接造訪 | 140 | 129.64萬 |
| 使用者評分 | 4.3 | 4.5 |
| 支援平台 | 網站 | 網站 |

用你自己的美術素材訓練 AI 模型,在一條管線裡完成圖像、影片、3D 與音訊生產
用 10–30 張參考圖訓練風格模型,或用 5–15 張訓練特定角色/物件(1024×1024 效果最佳),讓新資產始終貼合品牌調性。訓練通常在 30 分鐘到數小時內完成,Multi-LoRA 合併還能在一次生成中同時組合風格與主體。
匯集 50 多家供應商的圖像、影片、3D 與音訊模型,並隨發布持續更新,可在同一需求下橫向比較。在工作流中替換模型時,參考圖與提示詞保持不變。
在視覺化畫布上串接模型與工具,把鏈路發布成全團隊一鍵執行的應用;也可讓 Node Agent 依文字需求自動生成節點圖,或由 Genie 填好表單。執行前會顯示所需點數。
透過 REST API 或 TypeScript、Python SDK 在自有管線中呼叫任意模型或已儲存工作流;透過 Scenario MCP 伺服器連接 Claude Code、Codex 或 Cursor;或用實驗性外掛在 Blender 內生成貼圖、道具與場景。
WaveSpeedAI is an AI media generation platform that exposes 1,000+ image, video, audio, 3D and language models through a single unified API. Developers call any model from Node, Python or cURL, while creators use hosted generators, Cinema Studio, a desktop app and CLI/MCP tooling to produce images, cinematic video, speech and music without managing infrastructure.
One API key and one SDK reach text-to-image, image editing, text-to-video, image-to-video, avatar, speech, music, 3D and LLM models, so you stop stitching together vendors.
Optimised GPU clusters aim for sub-second image generation and up to 4x faster video rendering, with a claimed 99.99% uptime.
Direct cinematic shots from a prompt with camera moves, film looks by genre and era, colour palettes and lighting controls.
Run any supported model from the terminal or from an AI agent, and use a desktop app when you would rather click than code.

專項更強提供注重隱私的開發者 API 與開源模型,但比起擴展媒體處理流程,它更優先支援無限制的文字生成。
專項更強為產製級創意素材打造的 AI 圖像與影片生成
專項更強結合生成式 AI 影片與圖片工具及開發者 API,但整體架構更偏向創意工作室應用程式。
專項更強結合生成式 AI 影片與圖片工具及開發者 API,但整體架構更偏向創意工作室應用程式。
Recraft is an AI design platform built around its own image models, generating and editing raster images and true editable vector graphics. It combines a text-to-image generator with a vector generator, background remover, image upscaler, mockup generator, and photo editor, plus reusable custom styles and team workspaces.
Recraft is an AI design platform built around its own image models, generating and editing raster images and true editable vector graphics. It combines a text-to-image generator with a vector generator, background remover, image upscaler, mockup generator, and photo editor, plus reusable custom styles and team workspaces.
提供多模型整合 API 閘道器,但比起媒體生成,更專注於 LLM 路由與計費管理。