
Velo is an AI video platform for work that turns company context into clear, reliable videos. It covers demos, explainers, training, onboarding, support replies, and agent-generated videos, and gives teams one system to create, edit, update, translate, host, and measure video output.
| 面向/指標 | ![]() | |
|---|---|---|
| 官方網域 | deepseek.com | usevelo.ai |
| 定價模式 | 免費增值 | 免費增值 |
| 每月造訪次數 | 3.73億 (372,761,591) | 2.76萬 (27,556) |
| 月流量變化量 | +1536.22萬 (+4.3%) | -750 (-2.6%) |
| 主要市場與市佔 | 🇨🇳 中國 (50.5%) | 🇺🇸 美國 (22.6%) |
| 每月直接造訪 | 2.92億 | 1.07萬 |
| 使用者評分 | 4.8 | 4.2 |
| 支援平台 | 網站 | 網站 |
由 DeepSeek-V4 與 R1 驅動的高效開源權重與託管推理 AI,token 成本極低。
依託稀疏混合專家架構(V4-Pro 為 1.6T 中的 49B 啟用參數),以極低延遲與記憶體開銷處理最多 100 萬 token。
由 DeepSeek-R1 開創並融入 V3.2 與 V4 思考模式,對複雜數學、邏輯與程式碼除錯展示逐步自驗證過程。
快取命中時,輸入 token 成本降至 Flash 的每百萬 0.003 美元與 V4-Pro 的每百萬 0.022 美元,離峰 UTC 時段費率再低 50%。
在 chat.deepseek.com 及 iOS/Android 上免費提供文件上傳、推理模式與網頁搜尋,核心模型無須付費牆。

Velo is an AI video platform for work that turns company context into clear, reliable videos. It covers demos, explainers, training, onboarding, support replies, and agent-generated videos, and gives teams one system to create, edit, update, translate, host, and measure video output.
Turn documents, knowledge bases, help center articles, and product context into narration-ready videos with an AI voice.
Start from an idea and let Velo's agent produce videos grounded in company context via knowledge bases, connectors, and MCPs.
Keep videos consistent with logos, colors, fonts, and voice by configuring a brand kit once, then translate videos into multiple languages in one click.
Create the video you envision by showing Velo's agent what to explain, without writing the perfect prompt, and use polished screen recording and an AI editor.

提供巨大的上下文視窗與深度的 Google 生態系整合,但對於純推理任務感覺較不夠專精。
提供巨大的上下文視窗與深度的 Google 生態系整合,但對於純推理任務感覺較不夠專精。

免費 / 開源提供同等級的開源權重基礎模型與多語言效能,但需要自行架構主機或第三方推論設定。
免費 / 開源Enables running open-source models locally on consumer hardware, but lacks DeepSeek's cloud-scale hosted reasoning power.
免費 / 開源Enables running open-source models locally on consumer hardware, but lacks DeepSeek's cloud-scale hosted reasoning power.
Vidu is a multimodal generative video platform from ShengShu Technology. It turns text prompts, still images, and reference footage into short cinematic clips, with tools for subject-consistent characters, anime styles, and sound-effect generation.
Vidu is a multimodal generative video platform from ShengShu Technology. It turns text prompts, still images, and reference footage into short cinematic clips, with tools for subject-consistent characters, anime styles, and sound-effect generation.
免費 AI 影片生成器,支援脣形同步與多語言配音