YTGET

YouTube · Facebook · X · TikTok · Instagram · Threads

Kimi K3.1 與 Opus 5.5 八項實測對比:建造速度與穩定度差異

YouTube 影片筆記

Kimi K3.1 與 Opus 5.5 八項實測對比:建造速度與穩定度差異

原片標題:I Tested China's New Model vs Claude Opus 5.5 (Kimi K3.1)

原片說明
🔥 Kimi K3.1 Pro here + benchmarks: https://www.kimi.ai/code?utm_campaign=TR_idCFfXg7&utm_content=&utm_medium=Youtube&utm_source=CH_kEMBez3l&utm_term= 💎 Get all course resources 100% free: https://skool.com/maker-zero/about ➡️ If this is your first video: I'm Nick Saraev. I run LeftClick, an AI growth agency. We've done work for large brands you're likely familiar with, like Mr. Beast, Anthropic, OpenAI, and a few others. In plain English: I help people build systems that generate leads, close deals, and scale businesses, mostly using new AI models like Claude. I also make videos about how to do all of that here on YouTube. It hasn't always been this way. My family immigrated from a Soviet bloc country (People's Republic of Bulgaria) in the 90s. They worked extremely hard, and we were broke for a very long time. I grew up wishing for more security, stability, and freedom over my own life. I don't have a programming degree, and never took formal education for IT or workflows. I learned it all myself from YouTube videos like you're watching now. As for my path: I started my first business at 19 in college. We threw events (mostly parties) at campus bars, and charged people for entry. That failed to provide me enough money to pay my bills, so at 21, I started another business, a door-to-door marketing agency. From there, every year, I would start several companies, over 10 in total. I did this in a variety of industries, like e-commerce, photography, videography, SaaS, credit repair, etc. Each grew slightly more profitable, until, at 25, I would start 1SecondCopy, a content writing company. A friend and I eventually scaled that to over $90,000/m. It was my first taste of financial independence, and the reason I'm here today. Since starting my journey, I've grown a fair bit of capital. To diversify my income, at 30, I began investing in other businesses and advising companies on how to grow in the AI era. These include SaaS businesses, like Clairvo, agencies like Dental Connect, and more. Collectively, the businesses that I own outright or are a part of generate over $10 million per year. This returns far more than the 5-10% you would get in the stock market, but also requires a lot of my active time. There's a tradeoff there, but I greatly prefer active investing, where I have control over the entities I'm a part of. It wasn't easy to get here. But I did in the end! For anyone on their own entrepreneurial journey: I wish you luck, happiness, and perseverance. You can make it! Summary ⤵️ Kimi K3.1 Pro High is a great model. At a fraction of the tokens (and cost *per* token) of Opus-5.5, there’s a lot you can do with it. In this comprehensive test, I broke it down on a variety of UX, demo, design, and game tasks. Enjoy! Chapters 0:00 Kimi K3.1 Pro just dropped 0:44 Benchmarks 2:11 How I tested both models 2:45 Build time: Kimi vs Opus 3:03 3D keyboard configurator 5:02 Galaxy collision 6:27 Scroll-driven landing page 7:26 Tower demolition 8:16 Minecraft 9:22 Titancraft 10:15 Inside the Kimi app 11:03 Why Kimi's reasoning is so short 11:57 Kimi Swarm 12:49 Verdict: is Opus worth double the price?

Kimi K3.1 Pro 用 29 分鐘蓋完的東西,Opus 5.5 花了 47 分鐘還開不起來,這也太尷尬了吧。Nick Saraev 把兩個模型丟進同一組八個測試裡對打,從 3D 鍵盤、星系碰撞到 Minecraft,結果不是單純的「誰比較強」,而是誰比較不會出包。

Nick 把兩邊的測試成果放上同一個頁面
Nick 把兩邊的測試成果放上同一個頁面02:03

影片時間軸

  1. 00:44先掃過 Kimi 官網的跑分頁,Terminal Bench 4.0 拿下 65.7
  2. 01:10SWE Marathon 這種長時間工程任務,Kimi 反而贏過 Opus 5.5
  3. 02:34公佈平均建造時間,Kimi 29 分 41 秒,Opus 47 分 2 秒
  4. 02:52第一個測試是 3D 機械鍵盤,Kimi 能打字、能拆解、能加購物車
  5. 03:41換 Opus 5.5 上場,頁面根本打不開,Nick 只好叫它自己修
  6. 04:51星系碰撞模擬,Kimi 版被說看起來有點幼稚
  7. 07:14廢棄塔爆破模擬,Kimi 的城市背景空得可以
  8. 10:06進到 Kimi 自家的 Kimmy.ai 介面,示範 swarm 一次開一堆代理
Kestrel 塔從底部炸開,玻璃往下掉
Kestrel 塔從底部炸開,玻璃往下掉07:14

跑分贏了,但這不是重點

Nick 開頭就講明,跑分只能看出這麼多。他還是快速帶過一輪:Terminal Bench 4.0 拿 65.7,SWE Marathon 這種連續多小時的工程任務甚至贏過 Opus 5.5 快 5 分,Humanity's Last Exam 也站上 70。數字漂亮歸漂亮,他沒有停下來吹,馬上就切進實測,這個節奏我喜歡。

SWE Marathon 的長時程任務分數比較
SWE Marathon 的長時程任務分數比較01:10

八個測試裡 Opus 摔了兩次

鍵盤測試是整支影片最好看的一段。Kimi 版不只能換鍵帽顏色、換軸體,還能真的打字、把整把鍵盤炸開來看結構,Nick 自己都說比預期高級。反觀 Opus 5.5 頁面直接掛掉,得回頭叫它修。後面 Minecraft 那關更慘,Opus 版畫面漂亮到不行,但角色根本動不了,Nick 只能苦笑說它又搞砸一次。

Kimi 版 Minecraft,關燈後水面開始模糊
Kimi 版 Minecraft,關燈後水面開始模糊08:20

美感這回事,Kimi 真的輸

星系碰撞那段,Kimi 的版本被形容成有點幼稚,Opus 的星空明顯更有太空感。深海潛水網站也是,Opus 多做了一隻遊動的鯊魚和 3D 效果, scroll 起來好玩很多,代價是頁面變慢。Nick 自己承認這比較偏品味問題,不完全是設計品質,但差距擺在那裡。

Abyssal 潛水網站,Opus 版多了鯊魚
Abyssal 潛水網站,Opus 版多了鯊魚06:47

便宜一半,值得嗎

Kimi 每百萬輸入 token 兩塊美金、輸出十塊,Opus 5.5 直接翻倍。Nick 的結論很實在:Kimi 在很多地方沒 Opus 好,但穩定得多,該做的東西都做得出來,不會中途爛掉。他補一句,這些都是在方案額度裡跑的,不是 API,而 Kimi 給的額度相當寬鬆。

值得下載嗎

值得,如果你要的是「便宜、穩定、不會開天窗」的 coding 代理,這支把八個測試的失敗畫面都留著給你看,比看跑分表實在。

覺得這篇筆記有幫助嗎?
YTGet 編輯室|看完影片整理的筆記,細節以原片為準。

連線中

影片封面

擷取畫面

輸入時間點,例如 3:20、3.20 或 200

影片皆已含音訊。檔案越大準備時間越久,完成前請保持頁面開啟。