跳到主内容
@wquguru
精选88Simon Willison 博客(RSS)模型发布/更新多源精选 ×8

Moonshot发布Kimi K3.1 Pro API

Introducing Kimi K3.1 Pro

原文
发到 X
推荐理由

新旗舰模型发布,定价策略清晰且极具竞争力,Benchmark数据详实,对API调用者有直接参考价值。

Introducing Kimi K3.1 Pro

介绍 Kimi K3.1 Pro

Kimi K3.1 Pro landed on the Kimi API today - it's not yet open weights, but Moonshot promise "Full model weights will be released by Nov 4, 2026".

Kimi K3.1 Pro 今日上线 Kimi API——目前尚未开放权重,但 Moonshot 承诺“完整模型权重将于 2026 年 11 月 4 日前发布”。

They're also offering a "Highspeed" version that's 2-3x faster, at twice the price (except for input token cache hits, which are priced the same).

他们还提供了一个“高速”版本,速度提升 2-3 倍,价格翻倍(输入 token 缓存命中除外,其定价相同)。

The US dollar pricing is $2.00/million tokens for input and $10/million for output - the same as GPT-6.1 Sol and half the price of Claude Opus 5.5.

美元定价为:输入每百万 token 2.00 美元,输出每百万 token 10 美元——与 GPT-6.1 Sol 相同,是 Claude Opus 5.5 价格的一半。

Moonshot's own benchmarks show Kimi as approximately level with Opus 5.5 and GPT-6 Astra - beating them on a few benchmarks, coming close on others.

Moonshot 自身的基准测试显示,Kimi 的性能大致与 Opus 5.5 和 GPT-6 Astra 持平——在部分基准测试中胜出,在其他测试中也接近。

Artificial Analysis scored Kimi at 55 - ahead of Claude Fable 5.1, GPT-6 Astra, and Gemini 4 Argon (all at 53) and behind only Claude Opus 5.5 (58).

Artificial Analysis 给 Kimi 打分为 55 分——高于 Claude Fable 5.1、GPT-6 Astra 和 Gemini 4 Argon(均为 53 分),仅低于 Claude Opus 5.5(58 分)。

This is the first time a (soon-to-be) open weight model has scored this highly against the frontier.

这是首次有(即将)开源权重的模型在与前沿模型的对比中获得如此高的评分。

I tried it out via the Moonshot API using the llm openai endpoint command I added to LLM 0.32 back in August:

我通过 Moonshot API 尝试了该模型,使用的是我在今年 8 月添加到 LLM 0.32 中的 llm openai endpoint 命令:

代码 · 9 行
llm openai endpoint \
  https://api.moonshot.ai/v1 \
  --key kimi \
  -m kimi-k3.1-pro \
  "Generate an SVG of a pelican riding a bicycle" \
  --responses \
  -o reasoning_effort low \
  > kimi-k3.1-pro-pelican-low.stdout.txt \
  2> kimi-k3.1-pro-pelican-low.stderr.txt

llm openai endpoint can run prompts against any OpenAI-compatible endpoint but doesn't log the results, hence the extra options to save stdout and stderr (the reasoning trace).

llm openai endpoint 可以对任何兼容 OpenAI 的端点运行提示词,但不记录结果,因此需要额外选项来保存 stdout 和 stderr(推理轨迹)。

Here are the pelicans I got for Kimi K3.1 Pro on the three reasoning levels it supports - low, medium, and max. This is max:

以下是我在 Kimi K3.1 Pro 支持的三个推理级别(低、中、高)下得到的鹈鹕图像——这是最高级别的结果:

For comparison, I ran fresh pelicans against Kimi K3 on those three reasoning levels.

作为对比,我在同样的三个推理级别下对 Kimi K3 运行了新的鹈鹕图像生成。

Via Hacker News

来源:Hacker News

Tags: ai, generative-ai, llms, llm, llm-pricing, pelican-riding-a-bicycle, llm-release, ai-in-china, moonshot, kimi

标签:ai, generative-ai, llms, llm, llm-pricing, pelican-riding-a-bicycle, llm-release, ai-in-china, moonshot, kimi

更进一步:量化金融体系

看懂新闻只是起点——沿量化金融路径,把它变成能交付的工程能力

进入量化体系 →