AI News
insight Simon Willison 2026-10-13T22:28 正文 ok 266 words

llm-gemini 0.35 发布：新增 Gemini 4 Argon 支持

英文标题：llm-gemini 0.35

llm-gemini 0.35 新增对 Gemini 4 Argon 的支持，该模型现已通过 Gemini API 和每月 100 美元订阅计划向用户提供。作者测试了模型生成的低、中、高复杂度鹈鹕 SVG，认为结果均十分出色。Gemini 4 Argon 首发价格为每百万输入 token 2 美元、每百万输出 token 10 美元，但计划于 12 月 1 日涨价一倍。Artificial Analysis 的基准数据显示，Argon 完成整套测试的总成本为 2407 美元，显著低于 OpenAI、Anthropic 和 Meta 的最新模型，对关注 Agent 实际任务成本的开发者具有参考价值。
打开原文相关 6 · 机会 2
media56.1 KB media138.4 KB

中文

我为 llm-gemini 新增了对 Gemini 4 Argon 的支持。这款新模型于今天通过 Gemini API 向外提供，也面向其每月 100 美元订阅计划的客户开放。

这里是低、中、高三个版本的鹈鹕。不出所料，它们都非常出色：Gemini 一直以来都很擅长生成 SVG，而且据我所知，Google 是唯一一家其员工曾专门投入 SVG 研究的实验室（他们此前也曾炫耀过自家的鹈鹕）。

下面是“高”版本的插图，成本为 17.166 美分：

时尚而准确的自行车车架。篮子里有鱼。漂亮的围巾。双脚分别位于车架两侧并踩在踏板上。带辐条、一个小帽子、漂亮的喙和喉囊，以及略显凶狠的表情。

定价很有意思。Argon 在发布时的价格是每百万输入 token 2 美元、每百万输出 token 10 美元——与 GPT-6 Sol 价格相同，是 Claude Opus 5.5 价格的一半，也是 Claude Fable 5.1 和 GPT-6 Astra 价格的五分之一。

但 Argon 计划于 12 月 1 日价格翻倍——届时将与 Opus 5.5 同价。

如今，这类价格其实已经不太能直接比较，因为真正重要的是：模型在 Agent 执行框架中为达成目标需要消耗多少 token。

一个有用的线索来自 Artificial Analysis，他们发布了使用每个模型运行其基准测试的总成本。Argon 在这项指标上的表现相当好：运行该基准测试的成本为 2,407 美元，接近 GLM-5.3，且显著低于 OpenAI、Anthropic 和 Meta 的最新模型。

即使价格翻倍后，它仍会比 GPT-6 Astra 略便宜。

Artificial Analysis 堆叠条形图截图，比较了 14 个 AI 模型的美元成本，按从低到高排序。条形按提供商着色，每个条形下方有提供商徽标，底部标注倾斜的模型名称。右上角徽标：Artificial Analysis。指向 Gemini 4 Argon 条形的悬浮提示框显示：Gemini 4 Argon (High)，$2,407，输出 $313，推理 $815，缓存写入 $897，缓存读取 $264，非缓存输入 $118，点击获取更多信息。从左至右的条形及其上方总成本、部分条形内部标注的分项数值为：Qwen3.8 27B (xhigh)，$1,336，Alibaba 徽标；Mistral Large 4 Preview，$1,607，Mistral 徽标；Gemini 3.8 Flash (high)，$1,623，Google 徽标；Muse Spark 1.3 (max)，$2,000，分项 $769，Meta 徽标；Gemini 4 Argon (high)，$2,407，分项 $815 和 $897，Google 徽标；GLM-5.3 (max)，$2,503，分项 $790，Z.ai 徽标；Kimi K3 (max)，$3,658，Kimi 徽标；Qwen3.8 Max (0902)，$4,935，分项 $893 和 $1,441，Alibaba 徽标；Grok 4.7 (xhigh)，$4,967，分项 $1,861 和 $1,618，xAI 徽标；GPT-6 Astra (max)，$5,324，分项 $778 和 $978，OpenAI 徽标；Muse Wonder (max)，$5,733，分项 $1,748，Meta 徽标；Claude Sonnet 5.5 (max with fallback)，$7,260，分项 $1,216 和 $1,815，Anthropic 徽标；Claude Opus 5.5 (max with fallback)，$8,708，分项 $921，Anthropic 徽标；Claude Fable 5.1 (max with fallback)，$13,129，分项 $7,219，Anthropic 徽标。

（我用 ChatGPT 编辑了这张图表！所有数据都正确，我已核对过——我需要让 Gemini 4 Argon 的信息框显示出来，同时又不与其他数字重叠。）

English / 原文

I added support to llm-gemini for the new Gemini 4 Argon model, made available today via the Gemini API and for customers of their $100/month subscription plan.

Here are the low, medium and high pelicans. As expected, they are all excellent: Gemini has a track record of producing great SVGs, and are the only lab where I know their staff have invested in SVG specifically (and who have boasted about their pelicans before.)

Here's the "high" illustration, which cost me 17.166 cents:

Stylish, accurate bicycle frame. Fish in the basket. Snazzy scarf. Feet on either side of the frame touching the pedals. Spokes, a little hat, a good beak and pouch, a somewhat fierce expression.

The pricing is interesting. At launch, Argon is $2/million input and $10/million output - the same price as GPT-6 Sol, half the price of Claude Opus 5.5 and 1/5th the price of Claude Fable 5.1 and GPT-6 Astra.

But on December 1st Argon is set to double in price - so it will be the same price as Opus 5.5.

Prices like this aren't really comparable these days since what really matters is how many tokens a model takes to achieve a goal within an agentic harness.

One useful clue for that comes from Artificial Analysis, who publish the total cost to run their benchmark using each model. Argon is looking very good there - costing $2,407 to run the benchmark, putting it close to GLM-5.3 and significantly cheaper than OpenAI, Anthropic, and Meta's latest models.

When the price doubles it will still be a little cheaper than GPT-6 Astra.

Screenshot of a stacked bar chart from Artificial Analysis comparing dollar costs for 14 AI models, sorted from lowest to highest, with bars colored by provider, the provider's logo beneath each bar and angled model names along the bottom. Logo at top right: Artificial Analysis. A hover tooltip pointing at the Gemini 4 Argon bar reads: Gemini 4 Argon (High), $2,407, Output $313, Reasoning $815, Cache Write $897, Cache Read $264, Non-Cache Input $118, Click for more info. Bars from left to right, with the total above each bar and some segment values printed inside: Qwen3.8 27B (xhigh), $1,336, Alibaba logo; Mistral Large 4 Preview, $1,607, Mistral logo; Gemini 3.8 Flash (high), $1,623, Google logo; Muse Spark 1.3 (max), $2,000, segment $769, Meta logo; Gemini 4 Argon (high), $2,407, segments $815 and $897, Google logo; GLM-5.3 (max), $2,503, segment $790, Z.ai logo; Kimi K3 (max), $3,658, Kimi logo; Qwen3.8 Max (0902), $4,935, segments $893 and $1,441, Alibaba logo; Grok 4.7 (xhigh), $4,967, segments $1,861 and $1,618, xAI logo; GPT-6 Astra (max), $5,324, segments $778 and $978, OpenAI logo; Muse Wonder (max), $5,733, segment $1,748, Meta logo; Claude Sonnet 5.5 (max with fallback), $7,260, segments $1,216 and $1,815, Anthropic logo; Claude Opus 5.5 (max with fallback), $8,708, segment $921, Anthropic logo; Claude Fable 5.1 (max with fallback), $13,129, segment $7,219, Anthropic logo.

(I used ChatGPT to edit that chart! The data is all correct, I checked - I needed to get that Gemini 4 Argon info box to display without overlapping the other numbers.)