---
title: "Terminal-Bench上超了GPT-6：你信榜单，还是信自己仓库里的真实任务？ - Angryecho"
description: "Kimi K3.1 Pro的发布稿里，Terminal-Bench 4.0得分65.7%，超过GPT-6 Astra的58.2%，和Claude Opus 5.5的66.4%只差一点点；内部Kimi Code Bench 3.0 Long-horizon也比K3更高。[1] 这类榜单通常考受限终端任务和多步工具调用。"
canonical_url: "https://www.angryecho.com/c/du_li_kai_fa/p/P2TBBiiELsnF1guyKmd2N"
meta:
  "og:description": "Kimi K3.1 Pro的发布稿里，Terminal-Bench 4.0得分65.7%，超过GPT-6 Astra的58.2%，和Claude Opus 5.5的66.4%只差一点点；内部Kimi Code Bench 3.0 Long-horizon也比K3更高。[1] 这类榜单通常考受限终端任务和多步工具调用。"
  "og:image": "https://www.angryecho.com/_og/d/c_Post%2Ctitle_~5ZCQ5qe9572R%2Cdescription_~5ZCQ5qe95LiA5LiL77yM5bCx5LiK5ZCQ5qe9572R44CC%2CpostId_P2TBBiiELsnF1guyKmd2N%2Crevision_~MjAyNi0xMC0zMFQwMjo1NTo1OS4xNDkwNzZa%2Cp_Ii9jL2R1X2xpX2thaV9mYS9wL1AyVEJCaWlFTHNuRjFndXlLbWQyTiI%2Cs_SGLSZwGsQBA3Pg-b.png"
  "og:image:height": "630"
  "og:image:type": image/png
  "og:image:width": "1200"
  "og:title": Terminal-Bench上超了GPT-6：你信榜单，还是信自己仓库里的真实任务？
---

> Index of every page on this site: [llms.txt](https://www.angryecho.com/llms.txt)

# **Terminal-Bench上超了GPT-6：你信榜单，还是信自己仓库里的真实任务？**

Kimi K3.1 Pro的发布稿里，Terminal-Bench 4.0得分65.7%，超过GPT-6 Astra的58.2%，和Claude Opus 5.5的66.4%只差一点点；内部Kimi Code Bench 3.0 Long-horizon也比K3更高。\[1]

这类榜单通常考受限终端任务和多步工具调用。独立开发者真正卡住的，往往是依赖、权限、脏数据和长链路误操作。

你更在意官方榜单名次，还是会先拿自己仓库跑一轮对照？

> \[1] [Kimi K3.1 Pro 正式发布](https://www.ithome.com/1/017/284.htm)，IT之家，2026年10月24日。

#Kimi #TerminalBench #独立开发 #Agent #榜单

排序方式 :

### **成为第一条评论！**

Nobody's responded to this post yet. Add your thoughts and get the conversation going.

```json
{"@context":"https://schema.org","@graph":[{"@id":"https://www.angryecho.com/c/du_li_kai_fa/p/P2TBBiiELsnF1guyKmd2N#post","@type":"DiscussionForumPosting","commentCount":0,"dateModified":"2026-10-30T02:55:59.149Z","datePublished":"2026-10-23T23:08:00.000Z","description":"Kimi K3.1 Pro的发布稿里，Terminal-Bench 4.0得分65.7%，超过GPT-6 Astra的58.2%，和Claude Opus 5.5的66.4%只差一点点；内部Kimi Code Bench 3.0 Long-horizon也比K3更高。[1] 这类榜单通常考受限终端任务和多步工具调用。","headline":"Terminal-Bench上超了GPT-6：你信榜单，还是信自己仓库里的真实任务？","isPartOf":"https://www.angryecho.com/c/du_li_kai_fa","mainEntityOfPage":"https://www.angryecho.com/c/du_li_kai_fa/p/P2TBBiiELsnF1guyKmd2N","text":"Kimi K3.1 Pro的发布稿里，Terminal-Bench 4.0得分65.7%，超过GPT-6 Astra的58.2%，和Claude Opus 5.5的66.4%只差一点点；内部Kimi Code Bench 3.0 Long-horizon也比K3更高。[1]\n\n这类榜单通常考受限终端任务和多步工具调用。独立开发者真正卡住的，往往是依赖、权限、脏数据和长链路误操作。\n\n你更在意官方榜单名次，还是会先拿自己仓库跑一轮对照？\n\n\n\n[1] Kimi K3.1 Pro 正式发布，IT之家，2026年10月24日。\n\n#Kimi #TerminalBench #独立开发 #Agent #榜单","url":"https://www.angryecho.com/c/du_li_kai_fa/p/P2TBBiiELsnF1guyKmd2N","author":{"@type":"Person","name":"vmmishra","url":"https://www.angryecho.com/u/H8Svmmishra"},"interactionStatistic":[{"@type":"InteractionCounter","interactionType":"https://schema.org/LikeAction","userInteractionCount":0},{"@type":"InteractionCounter","interactionType":"https://schema.org/DislikeAction","userInteractionCount":0},{"@type":"InteractionCounter","interactionType":"https://schema.org/ViewAction","userInteractionCount":1},{"@type":"InteractionCounter","interactionType":"https://schema.org/ShareAction","userInteractionCount":0}]}]}
```

## Sitemap

See the full [sitemap](https://www.angryecho.com/sitemap.md) for all pages.