# Kimi K3: Complete A-Z Guide to the Open Model that Beat Fable 5
**作者**: Kirill
**日期**: 2026-07-16T18:58:25.000Z
**来源**: [https://x.com/kirillk_web3/status/2078880995705516361](https://x.com/kirillk_web3/status/2078880995705516361)
---
This is a complete A–Z breakdown of what Kimi K3 actually is, what it can do, and why it's quietly becoming the most important coding model nobody is talking about yet.
> Bookmark this page so you don't lose this article.
2.8 trillion parameters. 1 million token context.
Here's everything in it.
## Before We Talk About Benchmarks, Let's Talk About What Just Happened
For the last three years the story has been the same: American labs build the frontier, Chinese labs build the cheap copy six months later.
On July 16, Moonshot AI released Kimi K3.
> **Kimi.ai@Kimi_Moonshot**: [原文链接](https://x.com/Kimi_Moonshot/status/2077830229968683203)
>
> Introducing Kimi K3: Open Frontier Intelligence
> 2.8 Trillion Parameters, 1 Million Context, Native Multimodal
> Kimi Delta Attention enables up to 6.3x faster decoding in million-token contexts
> Attention Residuals deliver ~25% higher training efficiency at <2% additional
>
> 
> 
2.8 trillion total parameters — roughly 75% larger than DeepSeek's V4 Pro and nearly 4x Zhipu's GLM 5 series. The largest open-source model ever released.
It outperforms Claude Opus 4.8. It outperforms GPT-5.6 Sol. And on Moonshot's benchmarks, it sits on par with Fable 5 — the most capable model available to the public today.
The copy story is over.
## The Numbers
That last row is the one people are misreading. Let's deal with it properly.
## The Pricing Reality — Read This Carefully
Kimi K3 is not a discount model. At $3/$15 per million tokens, it's priced at roughly Claude Sonnet levels — not at the deep discounts Moonshot's earlier releases offered.
So where does "5x cheaper than Fable 5" come from?
Per-task cost, not per-token price.
K3 uses substantially fewer output tokens to complete the same work. On a representative coding task: Fable 5 costs around $1.30. K3 costs around $0.25.
Same output tier. Fewer tokens. Lower total bill.
The honest framing: Sonnet price, Fable-level experience. You're not buying the cheapest tokens on the market. You're buying frontier capability at mid-tier pricing, and getting there in fewer steps.
If you're comparing raw token prices, K3 looks unremarkable. If you're comparing what it costs to finish a job, it's a different conversation entirely.
## Where Kimi K3 Is State of the Art
According to Moonshot's published benchmarks, K3 sets SOTA on:
- HLE with tools — Humanity's Last Exam, tool-augmented
- SWE-Bench Pro — real-world software engineering
- SWE-Bench Multilingual — engineering across languages
- BrowseComp — web browsing and research
- Toolathlon — tool use at scale
- CharXiv with Python — chart and figure reasoning
- MathVision with Python — visual mathematics
Against the competition: outperforms Opus 4.8 and GPT-5.6 Sol, on par with Fable 5.
In blind testing by the AI evaluator Arena, developers preferred Kimi over every leading U.S. model.
## The Two Architecture Innovations That Actually Matter
This is the part most coverage is skipping, and it's the most interesting thing in the release.
- Kimi Delta Attention (KDA)
A hybrid linear attention mechanism. The result: up to 6.3x faster decoding in million-token contexts.
A 1M context window is only useful if you can actually work in it without waiting forever. KDA is what makes the context window practical instead of theoretical.
- Attention Residuals (AttnRes)
A drop-in replacement for residual connections. Delivers roughly 25% higher training efficiency at under 2% additional cost.
Both techniques were published as open research by the Moonshot team on GitHub before the model shipped. This isn't marketing — it's peer-inspectable work.
And combined, they explain the token efficiency: K3 uses 21% fewer output tokens than K2.6 on equivalent tasks.
## The Part That Should Make You Sit Up: K3 Optimized Its Own Architecture
Moonshot gave K3 the implementation of Attention Residuals — its own architectural component — and one objective: make it faster on H200 hardware without changing numerical behavior.
Then they left it alone.
20 hours of experiments. Zero human intervention. 1.6x speedup.
In a separate run, K3 cut the FLA Triton AttnRes forward/backward time from 283.6ms to 114.4ms — a 2.48x speedup — again without changing the numerics.
And it iterated faster than Fable 5 doing the same task.
Read that again. The model improved the mechanism that makes the model work. Autonomously. Overnight.
This is what "recursive self-improvement" looks like in practice — not a sci-fi scenario, a 20-hour job with a measurable result. It's the same capability Anthropic flagged as a reason for caution weeks ago, running in an open-weight model that anyone can download on July 27.
## Long-Horizon Coding Is the Whole Point
Moonshot's own framing: "K3 stands as Moonshot AI's most powerful open-source coding model to date. Operating with minimal human oversight, it can sustain long engineering sessions, navigate massive repositories, and orchestrate terminal tools."
The three things that matter for real engineering work:
1. Sustained sessions — the 20-hour autonomous experiment isn't a demo. It's the design target.
2. Massive repositories — a 1M context window with 6.3x faster decoding means you can put a real codebase in context and actually work in it.
3. Tool orchestration — SOTA on Toolathlon and BrowseComp means it handles the terminal, the browser, and the API calls without falling apart.
Frontend coding lands at Fable 5 level. AAA-quality game scenes from a single prompt. WebGPU, Three.js, shaders, physics — the stuff that used to require a studio.
## Vibe Coding With Kimi K3
This is where the model stops being a benchmark number and starts being obviously useful.
K3's frontend coding sits at Fable 5 level. In practice that means the wall between a description and a working 3D product is now one prompt.
What people have actually shipped from single prompts:
Games — WebGPU combat arenas with VFX. Browser-based 3D city traversal with grappling mechanics.
1. Zombie FPS.
2. Multiplayer racing.
3. A working 3D GBA emulator.
4. Full MMORPG scaffolding.
Not "a game-like demo" — playable scenes with hit reactions, impact feedback, and physics that behave.
*(视频内容)*
3D objects with real material work — A Rolex with correct glossiness and light behavior. A CSGO weapon that assembles and disassembles across 33 individually modeled parts. A Sony camera teardown. A black hole with gravitational lensing. Ocean simulation with actual wave dynamics.
*(视频内容)*
Physics scenes — Fabric simulation. Structural collapse. Vehicle collisions with momentum transfer that looks right instead of looking scripted.
The thing that makes this different from previous "AI builds a website" demos: K3 handles the details that normally break. Lighting. Shadows. Material glossiness. The 20 small things that separate "clearly AI-generated" from "someone built this."
One prompt. Four million tokens. A scene that used to need a team.
## 4 Battle-Tested Prompts (Copy-Paste Ready)
These are structured to exploit what K3 is actually good at: long-horizon work, tool use, and holding a large context without drifting.
Prompt 1 — The Physics Stress Test
This is the prompt that separates real capability from demo polish. If a model can do all three scenes at quality, it's real.
```
Build three self-contained HTML5 canvas scenes with real physics.
No external libraries. Everything in one file per scene.
Scene 1: A train derailing off a broken bridge into water.
Include: carriage momentum, bridge structural failure,
water displacement on impact.
Scene 2: Two cars jumping off ramps and colliding mid-air
over a canyon. Include: correct trajectory arcs,
collision momentum transfer, debris.
Scene 3: A monster truck crushing a row of parked cars.
Include: suspension compression, sheet metal deformation,
weight distribution.
Requirements for all three:
- Physics must be computed, not animated
- 60fps target
- Include a reset button
- Comment the physics math so I can verify it
Build all three. Don't ask clarifying questions.
```
Prompt 2 — The Long-Horizon Repo Task
This is what K3 was actually built for. Point it at a real codebase and give it an outcome, not a task.
```
Read this entire codebase before writing anything.
[paste your repo, or point it at the directory]
Objective: [state a measurable outcome — e.g. "reduce
p95 API response time by 30%" or "eliminate all N+1
queries in the data layer"]
Rules:
- Profile first. Show me where the time actually goes
before you change anything.
- Do not change public interfaces or numerical behavior.
- Run the existing test suite after each change.
- If a change makes things worse, revert it and tell me why.
- Work until the objective is hit or you can prove it's
not achievable without breaking the rules.
Report at the end: what you changed, what it bought,
what you tried that didn't work.
```
The "what you tried that didn't work" line matters. It's what turns the output from a diff into an engineering log.
Prompt 3 — The 3D Product Build
For anyone doing frontend, landing pages, or product visualization.
```
Build an interactive 3D [object] in the browser.
Single HTML file. Three.js.
The object: [describe it precisely — materials,
number of parts, what moves]
Must include:
- Correct material properties (glossiness, roughness,
metalness where relevant)
- Three-point lighting with real shadows
- Orbit controls
- [Any interaction — assembly/disassembly, animation,
hover states]
Quality bar: this should look like a product render,
not a WebGL demo. If a detail would be visible in real
life, model it.
Build the whole thing. Show me the file.
```
The "quality bar" line does more work than anything else in the prompt. K3 responds to standards, not just instructions.
Prompt 4 — Real-Time Feature With a Moving Part
Full-stack CRUD is one shape of app. Real-time is a different shape entirely — state has to stay in sync across clients, not just persist to a database. This is where a lot of models quietly fall apart.
```
Add a real-time feature to an existing app: a live collaborative
cursor + comment system, like Figma's.
REQUIREMENTS:
- WebSocket connection (use Socket.io or native ws)
- Show other connected users' cursor positions in real time
- Click anywhere to drop a comment pin
- Comments update live for all connected clients
- Handle disconnect/reconnect gracefully — no ghost cursors
- Debounce cursor updates so it doesn't flood the socket
BUILD:
1. Set up the WebSocket server
2. Build the client-side connection with auto-reconnect
3. Implement cursor broadcasting with debouncing
4. Implement the comment pin system
5. Add a simple presence indicator (who's online)
6. Test with at least 2 simulated clients to confirm sync works
Show me how you tested multi-client sync before calling this done.
```
Expected result: A working real-time layer in 15–30 minutes, with visible proof it was tested against more than one client.
The last line is the important part. Real-time bugs don't show up with one browser tab open — they show up with two. A model that skips multi-client testing will hand you code that looks done and isn't.
## When K3 Goes Wrong: Troubleshooting Guide
It's a new model with real rough edges. Here's what breaks and what to do.
- Problem: It burns tokens on trivial tasks
This is the big one, and it's structural.
K3 currently ships with only one reasoning effort level — 'max'. There's no "just answer quickly" mode. Independent testers clocked 13,241 reasoning tokens for generating a simple SVG — roughly $0.25 for something that should cost fractions of a cent.
> The fix: don't route simple work to K3. It's a frontier model with one gear, and that gear is "think hard about everything." Use K2.7 or a smaller model for boilerplate, formatting, and anything mechanical. Save K3 for work that justifies the reasoning.
This is the single biggest mistake people will make in the first month: making K3 the default for everything and then being surprised by the bill.
- Problem: The context window fills up anyway
1M tokens sounds infinite until you put a real repo in it and it isn't.
> The fix: don't dump everything. Point it at the directories that matter. K3's strength on long-horizon work comes from sustained focus, not from having every file in context. A tight 200K of the right code beats a bloated 900K of the whole monorepo.
- Problem: It drifts on very long autonomous runs
The 20-hour autonomous experiment is real, but it worked because the objective was measurable — "make this faster on H200 without changing numerics." Give it a vague goal and a long leash and it will wander.
> The fix: every long-horizon prompt needs a verifiable success condition. Not "improve the code." A number, a test that passes, a benchmark that moves. If you can't state how you'd check whether it succeeded, it will drift.
- Problem: You can't reproduce someone's benchmark result
Nobody outside Moonshot has audited this model. Weights don't drop until July 27. Every number circulating right now — including in this article — is vendor-reported.
> The fix: wait for July 27, or test on your own tasks and trust that instead. Your five real tasks are worth more than anyone's benchmark table.
- Problem: Integration errors after switching from OpenAI or Anthropic
K3 is OpenAI SDK compatible, which handles 90% of the migration. The remaining 10% is usually the reasoning behavior — K3 thinks by default and returns reasoning tokens you may not be expecting in your response handling.
> The fix: check your token accounting and your response parsing before you assume the model is broken. Most "K3 is expensive" complaints are actually "I forgot reasoning tokens are output tokens."
## The Context Nobody Should Ignore
Moonshot is based in Beijing, founded by Yang Zhilin, a former Google researcher, and backed by Alibaba.
> **Kirill@kirillk_web3**: [原文链接](https://x.com/kirillk_web3/status/2057528102368977328)
>
> instead of watching 2 hours of Netflix tonight, watch this 40-minute masterclass from the founder of a $20B China AI company
> it's the clearest explanation I've seen of how Agent Swarms and AI systems actually work at scale
> useful whether you've never built an agent in your life
>
> 
They built a 2.8-trillion-parameter frontier model under three years of escalating U.S. export controls on advanced chips.
Bank of America's analysts put it directly: "Despite persistent hardware/compute capacity constraints in China, K3 demonstrates that pre-training scaling, paired with architectural innovation, can still deliver step-change gains for flagship Chinese models."
And the timing isn't accidental: K3 landed days before the 2026 World Artificial Intelligence Conference in Shanghai.
## The July 27 Question
Weights drop July 27. That date matters more than the launch date.
Until then, K3 is a frontier model you can use through an API — impressive.
On July 27, it becomes something else: a frontier-class model that anyone can download, inspect, modify, and run on their own hardware. No API. No usage limits. No terms of service. No 30-day prompt retention.
That's the actual story. Not "China caught up."
China caught up and gave it away.
Worth noting: Chinese regulators have been discussing AI export controls of their own. Whether K3 is the last frontier open-weight release from China or the first of many is genuinely unclear right now.
## How to Try It
Chat: kimi.com — K3 is live now
API: OpenAI SDK compatible, so if you're already building on OpenAI or Anthropic toolchains, integration is a config change, not a rewrite
Weights: July 27
The SDK compatibility is a bigger deal than it sounds. Swapping models usually means rewriting your integration layer. Here it means changing a base URL and a model string.
## How to Install Kimi Code (The Terminal Agent)
If you want K3 running inside your actual codebase instead of a chat window, this is the setup.
Requirements:
- A computer (Mac, Windows, or Linux)
- Terminal access
- Kimi account — kimi.com
Step 1 — Install Kimi Code
Mac/Linux:
```
curl -fsSL https://code.kimi.com/kimi-code/install.sh | bash
```
Windows (PowerShell):
```
irm https://code.kimi.com/kimi-code/install.ps1 | iex
```
On Windows, install Git for Windows first — Kimi Code CLI uses its bundled Git Bash as the shell environment.
Verify the install:
```
kimi --version
```
> Due to macOS Gatekeeper, the first run may take noticeably longer. Add your terminal app in System Settings → Privacy & Security → Developer Tools to speed up later launches.
If you already have uv installed, you can install directly:
```
uv tool install --python 3.13 kimi-cli
```
Kimi Code CLI supports Python 3.12–3.14, with 3.13 recommended for best compatibility.
Step 2 — Navigate to your project and launch
```
cd your-project
kimi
```
On first launch, run /login inside the session to configure your API source — it opens a browser window for OAuth.
Step 3 — Give it a task
Kimi Code is now running inside your project, with direct read/write access to every file. Describe a task in plain English — it plans the steps, edits the code, runs tests, and reports what it did.
## What This Actually Means
If you're running production workloads:
Test it before switching. Vendor benchmarks and real-world performance diverge constantly. Pick five real tasks, run K3 and your current model side by side, measure cost per completed job — not cost per token.
The per-task math is the whole argument. At $3/$15 K3 isn't cheap on paper. At 21% fewer tokens and Fable-level output, it's cheap where it counts.
July 27 changes the calculus. Open weights means self-hosting, fine-tuning, and zero dependency on anyone's API staying online or anyone's terms staying favorable. For anything sensitive, that's not a small thing.
## Conclusion
The frontier used to be a place American labs built and everyone else visited six months later.
2.8 trillion parameters. 1M context. Fable 5-level coding at Sonnet pricing. Built under export controls, by the lowest-valued lab in the room, and given away for free on July 27.
The interesting question isn't whether K3 beats Fable 5 on some benchmark. It's what happens to the economics of closed frontier models when an open-weight one matches them.
## Links
- Kimi K3: https://www.kimi.com
- My Telegram: https://t.me/kirillk_web3
- My Twitter/X: https://x.com/kirillk_web3
- Partner Hosting: https://ishosting.com/affiliate/NzE0MiM2
Follow for more Vibe Coding information. Thank you for reading!
## 相关链接
- [Kirill](https://x.com/kirillk_web3)
- [@kirillk_web3](https://x.com/kirillk_web3)
- [337K](https://x.com/kirillk_web3/status/2078880995705516361/analytics)
- [Jul 17](https://x.com/Kimi_Moonshot/status/2077830229968683203)
- [23M](https://x.com/Kimi_Moonshot/status/2077830229968683203/analytics)
- [May 22](https://x.com/kirillk_web3/status/2057528102368977328)
- [13M](https://x.com/kirillk_web3/status/2057528102368977328/analytics)
- [kimi.com](https://kimi.com/)
- [kimi.com](https://kimi.com/)
- [https://www.kimi.com](https://www.kimi.com/)
- [https://t.me/kirillk_web3](https://t.me/kirillk_web3)
- [https://x.com/kirillk_web3](https://x.com/kirillk_web3)
- [https://ishosting.com/affiliate/NzE0MiM2](https://ishosting.com/affiliate/NzE0MiM2)
- [Upgrade to Premium](https://x.com/i/premium_sign_up)
- [12:33 AM · Jul 20, 2026](https://x.com/kirillk_web3/status/2078880995705516361)
- [337.8K Views](https://x.com/kirillk_web3/status/2078880995705516361/analytics)
- [View quotes](https://x.com/kirillk_web3/status/2078880995705516361/quotes)
---
*导出时间: 2026/7/21 09:41:46*
---
## 中文翻译
# Kimi K3：击败 Fable 5 的开放模型完整 A-Z 指南
**作者**: Kirill
**日期**: 2026-07-16T18:58:25.000Z
**来源**: [https://x.com/kirillk_web3/status/2078880995705516361](https://x.com/kirillk_web3/status/2078880995705516361)
---
这是一份完整的 A–Z 解析，说明 Kimi K3 到底是什么、它能做什么，以及为什么它正在悄无声息地成为最重要的、却还没人谈论的编程模型。
> 收藏本页，以免丢失这篇文章。
2.8 万亿参数。100 万 token 上下文。
以下是其中的全部内容。
## 在谈基准测试之前，先谈谈刚刚发生了什么
过去三年，故事一直是同样的剧本：美国实验室打造前沿模型，中国实验室六个月后造出廉价复制品。
7 月 16 日，月之暗面（Moonshot AI）发布了 Kimi K3。
> **Kimi.ai@Kimi_Moonshot**: [原文链接](https://x.com/Kimi_Moonshot/status/2077830229968683203)
>
> Kimi K3 正式发布：开放前沿智能
> 2.8 万亿参数、100 万上下文、原生多模态
> Kimi Delta Attention 在百万 token 上下文中实现最高 6.3 倍的解码加速
> Attention Residuals 以不到 2% 的额外成本带来约 25% 的训练效率提升
>
> 
> 
2.8 万亿总参数——大约比 DeepSeek 的 V4 Pro 大 75%，几乎是智谱 GLM 5 系列的 4 倍。有史以来发布的最大开源模型。
它超越了 Claude Opus 4.8。超越了 GPT-5.6 Sol。而在月之暗面自己的基准测试中，它与 Fable 5 不相上下——后者是当今公众可用的最强模型。
“复制品”的故事结束了。
## 数据
最后一行是人们容易误读的地方。让我们认真处理它。
## 定价实情——请仔细阅读
Kimi K3 不是打折模型。每百万 token $3/$15 的价格，大致相当于 Claude Sonnet 水平——并非月之暗面早期版本提供的深度折扣价。
那“比 Fable 5 便宜 5 倍”从何而来？
来自单任务成本，而非单 token 价格。
K3 完成同样的工作所用的输出 token 大幅更少。在一个代表性的编程任务上：Fable 5 大约花费 $1.30。K3 大约花费 $0.25。
同样的输出水平。更少的 token。更低的账单总额。
更客观的说法是：Sonnet 的价格，Fable 级别的体验。你买的不是市面上最便宜的 token。你以中端价格购买前沿能力，而且用更少的步骤达到目标。
如果你只比较原始 token 价格，K3 看起来平平无奇。但如果你比较的是完成一项工作的成本，那完全是另一回事了。
## Kimi K3 在哪些领域达到业界最先进水平
根据月之暗面公布的基准测试，K3 在以下领域创造了 SOTA：
- HLE with tools——人类最后的考试（工具增强版）
- SWE-Bench Pro——真实世界软件工程
- SWE-Bench Multilingual——跨语言工程
- BrowseComp——网页浏览与研究
- Toolathlon——大规模工具使用
- CharXiv with Python——图表与图形推理
- MathVision with Python——可视化数学
与竞争对手相比：超越 Opus 4.8 和 GPT-5.6 Sol，与 Fable 5 持平。
在 AI 评测机构 Arena 的盲测中，开发者在所有领先的美国模型中更偏爱 Kimi。
## 真正重要的两大架构创新
这是大多数报道都略过的部分，也是这次发布中最有意思的东西。
- Kimi Delta Attention（KDA）
一种混合线性注意力机制。结果：在百万 token 上下文中，解码速度最高提升 6.3 倍。
100 万的上下文窗口只有在你能在其中实际工作而无需漫长等待时才有用。KDA 正是让上下文窗口从理论变为实用的关键。
- Attention Residuals（AttnRes，注意力残差）
残差连接的即插即用替代方案。以不到 2% 的额外成本带来约 25% 的训练效率提升。
这两项技术都在模型发布之前，由月之暗面团队作为开放研究发布在 GitHub 上。这不是营销——这是可供同行审视的工作。
两者结合，解释了 token 效率：在同等任务上，K3 比 K2.6 少使用 21% 的输出 token。
## 最该让你警觉的部分：K3 优化了自己的架构
月之暗面向 K3 提供了 Attention Residuals——它自己的架构组件——的实现，以及一个目标：在不改变数值行为的前提下，让它在 H200 硬件上运行得更快。
然后他们放手不管了。
20 小时的实验。零人工干预。1.6 倍加速。
在另一次运行中，K3 将 FLA Triton AttnRes 的前向/反向传播时间从 283.6ms 降到 114.4ms——2.48 倍加速——同样没有改变数值。
而且它迭代得比执行同样任务的 Fable 5 更快。
再读一遍。这个模型改进了让这个模型得以运作的机制。自主完成。一夜之间。
这就是“递归自我改进”在实践中的样子——不是科幻场景，而是一项 20 小时、有可衡量结果的工作。这正是 Anthropic 几周前作为需要谨慎对待的理由而指出的那种能力，如今运行在一个 7 月 27 日任何人都可以下载的开放权重模型中。
## 长时程编程才是核心所在
月之暗面自己的表述：“K3 是月之暗面迄今为止最强大的开源编程模型。在极少人工监督的情况下运行，它能够支撑漫长的工程会话、在庞大的代码仓库中导航，并统筹各类终端工具。”
对真实工程工作至关重要的三件事：
1. 持续会话——20 小时的自主实验并非演示，而是设计目标。
2. 庞大代码仓库——100 万上下文窗口加上 6.3 倍解码加速，意味着你可以把一个真实的代码库放进上下文并真正在其中工作。
3. 工具统筹——在 Toolathlon 和 BrowseComp 上达到 SOTA，意味着它能够驾驭终端、浏览器和 API 调用而不会崩溃。
前端编码达到 Fable 5 水平。单个提示词即可生成 AAA 级游戏场景。WebGPU、Three.js、着色器、物理引擎——这些过去需要一整个工作室才能完成的东西。
## 用 Kimi K3 进行氛围编程（Vibe Coding）
这正是该模型不再只是一个跑分数字、开始明显好用的地方。
K3 的前端编码处于 Fable 5 水平。在实践中，这意味着从描述到一个可用的 3D 产品之间的鸿沟，如今只需一个提示词。
人们用单个提示词实际做出的东西：
游戏——带有特效的 WebGPU 战斗竞技场。基于浏览器的 3D 城市穿行，带抓钩机制。
1. 僵尸第一人称射击游戏。
2. 多人赛车游戏。
3. 一个可用的 3D GBA 模拟器。
4. 完整的 MMORPG 脚手架。
不是“类似游戏的演示”——而是带打击反应、冲击反馈和真实物理表现的可玩场景。
*(视频内容)*
材质表现出色的 3D 物体——一块光泽度和光影行为都正确的劳力士手表。一把可组装和拆解、由 33 个独立建模部件组成的 CSGO 武器。索尼相机的拆解展示。带引力透镜效应的黑洞。拥有真实波动动力学的海洋模拟。
*(视频内容)*
物理场景——布料模拟。结构坍塌。车辆碰撞，动量传递看起来真实自然，而不是像预设好的脚本。
这与以往“AI 搭建一个网站”演示的区别在于：K3 能处理好那些通常容易出错的地方。光照。阴影。材质光泽度。正是这 20 个小细节，把“明显是 AI 生成的”和“有人亲手做出来的”区分开来。
一个提示词。四百万 token。过去需要一个团队才能完成的场景。
## 4 个经过实战检验的提示词（可直接复制粘贴）
这些提示词经过精心设计，专门发挥 K3 真正擅长的方面：长周期任务、工具使用，以及在不跑偏的情况下保持大上下文。
提示词 1 —— 物理压力测试
这个提示词能区分真实能力与演示级的表面功夫。如果一个模型能以高质量完成全部三个场景，那就是真本事。
```
Build three self-contained HTML5 canvas scenes with real physics.
No external libraries. Everything in one file per scene.
Scene 1: A train derailing off a broken bridge into water.
Include: carriage momentum, bridge structural failure,
water displacement on impact.
Scene 2: Two cars jumping off ramps and colliding mid-air
over a canyon. Include: correct trajectory arcs,
collision momentum transfer, debris.
Scene 3: A monster truck crushing a row of parked cars.
Include: suspension compression, sheet metal deformation,
weight distribution.
Requirements for all three:
- Physics must be computed, not animated
- 60fps target
- Include a reset button
- Comment the physics math so I can verify it
Build all three. Don't ask clarifying questions.
```
提示词 2 —— 长周期代码库任务
这才是 K3 真正的用武之地。把它指向一个真实的代码库，给它一个目标结果，而不是一个具体任务。
```
Read this entire codebase before writing anything.
[paste your repo, or point it at the directory]
Objective: [state a measurable outcome — e.g. "reduce
p95 API response time by 30%" or "eliminate all N+1
queries in the data layer"]
Rules:
- Profile first. Show me where the time actually goes
before you change anything.
- Do not change public interfaces or numerical behavior.
- Run the existing test suite after each change.
- If a change makes things worse, revert it and tell me why.
- Work until the objective is hit or you can prove it's
not achievable without breaking the rules.
Report at the end: what you changed, what it bought,
what you tried that didn't work.
```
“你尝试过但没成功的方法”这一行很重要。它把输出从一段代码差异变成了一份工程日志。
提示词 3 —— 3D 产品构建
适合所有做前端、落地页或产品可视化的人。
```
Build an interactive 3D [object] in the browser.
Single HTML file. Three.js.
The object: [describe it precisely — materials,
number of parts, what moves]
Must include:
- Correct material properties (glossiness, roughness,
metalness where relevant)
- Three-point lighting with real shadows
- Orbit controls
- [Any interaction — assembly/disassembly, animation,
hover states]
Quality bar: this should look like a product render,
not a WebGL demo. If a detail would be visible in real
life, model it.
Build the whole thing. Show me the file.
```
“质量标准”这一行比提示词里其他任何内容都更有用。K3 对标准有响应，而不仅仅是对指令。
提示词 4 —— 带动态部件的实时功能
全栈 CRUD 是一种应用形态。实时应用完全是另一种形态——状态必须在多个客户端之间保持同步，而不仅仅是持久化到数据库。很多模型正是在这里悄悄垮掉的。
```
Add a real-time feature to an existing app: a live collaborative
cursor + comment system, like Figma's.
REQUIREMENTS:
- WebSocket connection (use Socket.io or native ws)
- Show other connected users' cursor positions in real time
- Click anywhere to drop a comment pin
- Comments update live for all connected clients
- Handle disconnect/reconnect gracefully — no ghost cursors
- Debounce cursor updates so it doesn't flood the socket
BUILD:
1. Set up the WebSocket server
2. Build the client-side connection with auto-reconnect
3. Implement cursor broadcasting with debouncing
4. Implement the comment pin system
5. Add a simple presence indicator (who's online)
6. Test with at least 2 simulated clients to confirm sync works
Show me how you tested multi-client sync before calling this done.
```
预期结果：在 15-30 分钟内得到一个可用的实时层，并有可见的证据证明它针对多个客户端进行过测试。
最后一行是最重要的。实时 bug 不会在只开一个浏览器标签页时出现——它们会在开两个的时候出现。一个跳过多客户端测试的模型，交给你的是看起来完成了、实际上没完成的代码。
## K3 出问题时的排错指南
这是一个仍有明显粗糙之处的新模型。以下是常见的故障点及应对方法。
- 问题：在琐碎任务上浪费 token
这是最大的问题，而且是结构性的。
K3 目前只提供一个推理强度级别——"max"。没有"快速回答"模式。独立测试人员测得，生成一个简单的 SVG 需要消耗 13,241 个推理 token——本应只需几分之一美分的事情，却花了大约 0.25 美元。
> 解决方法：不要把简单的工作交给 K3。它是一个只有一个挡位的前沿模型，而这个挡位就是"对所有事情都深思熟虑"。对于模板代码、格式化以及任何机械性的工作，请使用 K2.7 或更小的模型。把 K3 留给值得深度推理的工作。
这是人们在第一个月会犯的最大错误：把 K3 设为一切任务默认使用，然后对账单感到震惊。
- 问题：上下文窗口还是会被填满
100 万 token 听起来是无限的，直到你把一个真实的代码仓库放进去，才发现它不够用。
> 解决方法：不要一股脑全丢进去。把目标指向真正重要的目录。K3 在长周期任务上的优势来自持续的专注，而不是把每个文件都放进上下文。一个紧凑的 200K 正确代码，胜过臃肿的 900K 整个 monorepo。
- 问题：在超长的自主运行中会偏离目标
20 小时的自主实验是真实的，但它成功是因为目标可以衡量——"在 H200 上让它更快，同时不改变数值结果。"给它一个模糊的目标和很长的自由度，它就会游荡。
> 解决方法：每个长周期提示都需要一个可验证的成功条件。不要说"改进代码"。要给出一个数字、一个通过的测试、一个有变化的基准。如果你无法说明如何检验它是否成功，它就会偏离。
- 问题：你无法复现别人的基准测试结果
Moonshot 之外没有人审计过这个模型。权重要到 7 月 27 日才会发布。目前流传的每一个数字——包括本文中的——都是厂商自己报告的。
> 解决方法：等到 7 月 27 日，或者在你自己的任务上测试并以此为准。你的五个真实任务比任何人的基准测试表格都更有价值。
- 问题：从 OpenAI 或 Anthropic 切换过来后出现集成错误
K3 与 OpenAI SDK 兼容，这解决了 90% 的迁移工作。剩下 10% 通常是推理行为——K3 默认会思考，并返回推理 token，而你的响应处理逻辑可能没有预料到这些。
> 解决方法：在断定模型出问题之前，先检查你的 token 计量和响应解析逻辑。大多数"K3 太贵了"的抱怨，实际上是"我忘了推理 token 也算输出 token"。
## 任何人都不能忽视的背景信息
Moonshot 总部位于北京，由前谷歌研究员杨植麟创立，并获得阿里巴巴的支持。
> **Kirill@kirillk_web3**: [原文链接](https://x.com/kirillk_web3/status/2057528102368977328)
>
> 今晚别看 2 小时 Netflix 了，来看看这家估值 200 亿美元的中国 AI 公司创始人的 40 分钟大师课
> 这是我见过的对 Agent 集群（Agent Swarms）和 AI 系统如何大规模运作最清晰的讲解
> 无论你从未构建过 agent 还是已经经验丰富，都会有帮助
>
> 
在美国对先进芯片不断升级的出口管制之下，他们用了不到三年时间构建了一个 2.8 万亿参数的前沿模型。
美国银行的分析师直言："尽管中国在硬件/算力方面持续存在容量限制，但 K3 证明了预训练扩展与架构创新相结合，仍然能够为中国旗舰模型带来跨越式的提升。"
而发布时机并非偶然：K3 在上海 2026 世界人工智能大会开幕前几天落地。
## 7 月 27 日的问题
权重将于 7 月 27 日发布。这个日期比发布日期更重要。
在那之前，K3 是一个可以通过 API 使用的前沿模型——令人印象深刻。
7 月 27 日，它将变成另一个东西：一个任何人都可以下载、检视、修改并在自己硬件上运行的前沿级模型。没有 API 限制。没有用量限制。没有服务条款。没有 30 天的提示词保留期。
这才是真正的故事。不是"中国追赶上了"。
中国追赶上了，并且把它免费送了出去。
值得注意的是：中国的监管机构也一直在讨论自己的 AI 出口管制。K3 是中国最后一个前沿级开放权重模型，还是众多模型中的第一个，目前确实尚不明朗。
## 如何试用
聊天：kimi.com——K3 现已上线
API：与 OpenAI SDK 兼容，因此如果你已经在 OpenAI 或 Anthropic 工具链上构建，集成只需修改配置，而无需重写
权重：7 月 27 日
SDK 兼容性的意义比听起来更大。更换模型通常意味着重写你的集成层。而在这里，只需要改一个 base URL 和一个模型字符串。
## 如何安装 Kimi Code（终端 Agent）
如果你想让 K3 运行在你实际的代码库中，而不是聊天窗口里，那就是这个设置。
要求：
- 一台电脑（Mac、Windows 或 Linux）
- 终端访问权限
- Kimi 账号——kimi.com
第 1 步——安装 Kimi Code
Mac/Linux：
```
curl -fsSL https://code.kimi.com/kimi-code/install.sh | bash
```
Windows (PowerShell)：
```
irm https://code.kimi.com/kimi-code/install.ps1 | iex
```
在 Windows 上，请先安装 Git for Windows——Kimi Code CLI 使用其自带的 Git Bash 作为 shell 环境。
验证安装：
```
kimi --version
```
> 由于 macOS Gatekeeper 的原因，首次运行可能会明显更慢。在系统设置 → 隐私与安全性 → 开发者工具中添加你的终端应用，可以加快后续启动速度。
如果你已经安装了 uv，可以直接安装：
```
uv tool install --python 3.13 kimi-cli
```
Kimi Code CLI 支持 Python 3.12–3.14，推荐使用 3.13 以获得最佳兼容性。
第 2 步——进入你的项目并启动
```
cd your-project
kimi
```
首次启动时，在会话中运行 /login 来配置你的 API 源——它会打开一个浏览器窗口进行 OAuth 授权。
第 3 步——给它一个任务
Kimi Code 现在正在你的项目内运行，可以直接读写每一个文件。用通俗的英语描述一个任务——它会规划步骤、编辑代码、运行测试，并报告它做了什么。
## 这实际上意味着什么
如果你正在运行生产环境的工作负载：
切换之前先测试。厂商基准测试与现实性能经常不一致。挑五个真实任务，让 K3 和你当前使用的模型并排运行，衡量每个已完成任务的成本——而不是每 token 的成本。
按任务计算才是整个论证的核心。按 $3/$15 的定价，K3 在纸面上并不便宜。但考虑到 token 用量减少 21%，以及达到 Fable 级别的输出质量，它在真正重要的地方是便宜的。
7 月 27 日会改变整个格局。开放权重意味着可以自托管、微调，并且不必依赖任何人的 API 保持在线或任何人的条款保持有利。对于任何敏感的场景，这都不是小事。
## 结论
前沿曾经是美国实验室建立、其他人六个月后才到访的地方。
2.8 万亿参数。100 万上下文。以 Sonnet 的价格实现 Fable 5 级别的编程能力。在出口管制之下，由场内估值最低的实验室打造，并在 7 月 27 日免费开放。
有意思的问题不是 K3 能否在某些基准测试中击败 Fable 5，而是当一个开放权重模型追平它们时，闭源前沿模型的经济学会发生什么。
## 链接
- Kimi K3：https://www.kimi.com
- 我的 Telegram：https://t.me/kirillk_web3
- 我的 Twitter/X：https://x.com/kirillk_web3
- 合作伙伴托管：https://ishosting.com/affiliate/NzE0MiM2
关注获取更多 Vibe Coding 资讯。感谢阅读！
## 相关链接
- [Kirill](https://x.com/kirillk_web3)
- [@kirillk_web3](https://x.com/kirillk_web3)
- [337K](https://x.com/kirillk_web3/status/2078880995705516361/analytics)
- [Jul 17](https://x.com/Kimi_Moonshot/status/2077830229968683203)
- [23M](https://x.com/Kimi_Moonshot/status/2077830229968683203/analytics)
- [May 22](https://x.com/kirillk_web3/status/2057528102368977328)
- [13M](https://x.com/kirillk_web3/status/2057528102368977328/analytics)
- [kimi.com](https://kimi.com/)
- [kimi.com](https://kimi.com/)
- [https://www.kimi.com](https://www.kimi.com/)
- [https://t.me/kirillk_web3](https://t.me/kirillk_web3)
- [https://x.com/kirillk_web3](https://x.com/kirillk_web3)
- [https://ishosting.com/affiliate/NzE0MiM2](https://ishosting.com/affiliate/NzE0MiM2)
- [升级至 Premium](https://x.com/i/premium_sign_up)
- [2026年7月20日 凌晨12:33](https://x.com/kirillk_web3/status/2078880995705516361)
- [33.78万次浏览](https://x.com/kirillk_web3/status/2078880995705516361/analytics)
- [查看引用](https://x.com/kirillk_web3/status/2078880995705516361/quotes)
---
*导出时间：2026/7/21 09:41:46*
