AI News
insight Simon Willison 2026-10-07T20:56 正文 ok 548 words

Claude Haiku 5.5：Anthropic 推出快速低成本模型

英文标题：Claude Haiku 5.5

Anthropic 发布 Claude Haiku 5.5，将 10 万 token 以内的输入/输出价格降至每百万 token 0.10/0.50 美元，与 OpenAI GPT-6 Luna 持平，但超过 10 万 token 后价格将提高 5 倍。作者测试发现，Haiku 5.5 的新 tokenizer 对同一长提示词会多计约 1.25 倍 token，构成一项隐性涨价；在超长上下文场景下，Luna 的价格优势更明显。新模型默认启用中等推理强度，生成 SVG 鹈鹕插画的质量较前代明显提升。Anthropic 同时下调 Sonnet 5.5 缓存读取价格，并为 Max 和 Team 订阅用户提供每月 API 积分，降低订阅者使用 API 的门槛。
打开原文相关 2.4 · 机会 0
media35.6 KB media44.6 KB media51.9 KB media31.1 KB

中文

Claude Haiku 5.5

2026 年 10 月 7 日

正如此前承诺的那样，Anthropic 推出了新的快速、低成本模型：Introducing Claude Haiku 5.5。

上一代 Haiku 4.5 已经明显显得过时。它在近一年前发布，定价为每百万输入 token 1 美元、每百万输出 token 5 美元——即使在当时也相对昂贵，且是上个月发布的 OpenAI GPT-6 Luna价格的整整 10 倍。

新的 Haiku 在 100,000 token 以内的价格与 GPT-6 Luna 完全一致：0.10/0.50 美元。超过 100,000 token 后，其价格将提高 5 倍，达到 0.50/2.50 美元。Luna 本身也会在 272,000 token 时提价，但仅提高至 0.20/0.75 美元。

Haiku 5.5 还采用了新的、对用户不那么友好的 tokenizer。我的 Claude Token Counter 工具显示，与 Haiku 4.5 相比，同一份长提示词在 Haiku 5.5 中会消耗约 1.25 倍的 token，因此这里存在一项隐性的价格上涨。

如果你的工作负载能控制在 100,000 token 内，Haiku 的价格与 Luna 相同，且其公布的基准测试分数更高。但超过 100,000 token 后，Luna 看起来是划算得多的选择。

llm-anthropic 的最新版本终于修复了一个问题：我不再需要为每个新模型都发布该插件的新版本。我这样测试了新模型：

llm install -U llm-anthropic
llm anthropic refresh
llm -m claude-haiku-5.5 "Generate an SVG of a pelican riding a bicycle" -o thinking_effort low

鹈鹕

这里是低、中、高、xhigh 和 max 推理强度下生成的鹈鹕。新的 Haiku 不允许你关闭推理功能，默认值为 medium。除 low 外，其他所有强度下我都得到了不错的自行车车架。这个低推理强度的鹈鹕成本为 0.0936 美分，耗时 7 秒。

这只 max 推理强度的鹈鹕花了 5 分 9 秒才生成，但成本仍仅为 3.3826 美分：

扁平矢量插画：一只戴红帽的白鹈鹕从左向右骑着深灰色自行车穿过一片绿色草地。它修长的橙色双腿伸向橙色脚踏板，一只灰色翅膀向前伸出，握住弯曲的车把；巨大的黄橙色鸟喙及鼓起的橙色喉囊指向前方。身后白色速度线表现出运动感；上方的渐变蓝色天空中，有一轮带浅色光晕的黄太阳和两朵白云。

作为对比，这是我在一年前从 Haiku 4.5 获得的鹈鹕，成本为 0.7583 美分——Haiku 4.5 不支持推理强度等级。它画鹈鹕实在很糟：

由 Haiku 4.5 描述：一幅异想天开的插画，一只圆滚滚的棕褐色身体、粉色鸟喙和橙色双腿的鸟，在蓝天和绿草背景前骑着自行车。

面向订阅用户的慷慨 API 积分方案

除了 Haiku 5.5，Anthropic 今天还宣布将 Sonnet 5.5 的缓存读取价格减半。他们也为订阅计划增加了 API 积分：

第二，本周我们将为所有 Max 和 Team 订阅用户推出一项新的每月 API 积分，可在 Claude Platform 上使用。Max 5x 用户每月将获得 100 美元积分，Max 20x 用户将获得 200 美元，Team 订阅用户则最多可获得 500 美元，积分将在其用户之间共享。

领取方式相当简单：前往 Settings -> Billing，然后选择每月应获得这些积分的 API 组织：

Settings 对话框截图：右上角有关闭 X 按钮，顶部是一排标签，包括“General”“Account”“Privacy”“Billing”（已选中）“Usage”“Capabilities”“Memory”以及在边缘处被截断的“Design sys”。在一个带圆形花蕾的分枝植物线条图标旁，面板显示“Max plan”“20x more usage than Pro”以及“Your subscription will auto renew on Nov 2, 2026.”，右侧有“Adjust plan”按钮。下方是标题为“API credits”的区域，带蓝色“New”标记，内容为“Your Max 20x plan includes $200 USD in API credits each month.”；随后灰色文字写着“Link an API organization to start receiving credits.”，右侧有“Link organization”按钮。

API 积分金额恰好与订阅本身的费用相等。这确实非常慷慨——它让订阅用户更容易使用 API。Anthropic 还允许你关闭 API 的自动充值，代价是“当你的余额耗尽时，API 请求将停止”——如果你计划用完这些 API 积分，又不想冒着收到高额意外账单的风险，这正是你想要的功能。

请注意，每月积分不会结转——不用就会失效。

OpenAI 仍允许你将 Codex 订阅用于个人 API 使用；对于重度 API 用户而言，这仍然是更划算的方案。这项新的积分计划至少在一定程度上缩小了这一差距。

English / 原文

Claude Haiku 5.5

7th October 2026

As previously promised, here’s Anthropic’s new fast, low cost model: Introducing Claude Haiku 5.5.

The previous Haiku, 4.5, was very much showing its age. It came out almost a year ago, and was priced at $1/million input and $5/million output—relatively expensive even back then, and a full 10x the price of OpenAI’s GPT-6 Luna, released last month.

The new Haiku exactly matches the price of GPT-6 Luna—$0.10/$0.50—up to 100,000 tokens. Beyond 100,000 tokens the price increases 5x to $0.50/$2.50. Luna itself has a price increase at 272,000 tokens but only to $0.20/$0.75.

Haiku 5.5 also uses a new, less generous tokenizer. My Claude Token Counter tool shows that the same long prompt uses around 1.25x as many tokens with Haiku 5.5 compared to Haiku 4.5, so there’s a hidden price increase there.

If your workloads fit in 100,000 tokens, Haiku is the same price as Luna and reports higher benchmark scores. Above 100,000 tokens, Luna looks like a much better deal.

The most recent release of llm-anthropic finally fixed it so I don’t need to ship a new version of that plugin for every new model. I tested the new model like this:

llm install -U llm-anthropic
llm anthropic refresh
llm -m claude-haiku-5.5 "Generate an SVG of a pelican riding a bicycle" -o thinking_effort low

Pelicans

Here are pelicans for low, medium, high, xhigh, and max. The new Haiku doesn’t let you disable reasoning, and defaults to medium. I got a good bicycle frame for everything beyond low. The low effort pelican cost 0.0936 cents and took 7 seconds.

This max effort pelican took 5 minutes 9 seconds to generate, but still only cost me 3.3826 cents:

Flat vector illustration: a white pelican wearing a red cap rides a dark grey bicycle from left to right across a green grassy strip, its long orange legs reaching down to the orange pedals, a grey wing stretched forward to grip the curved handlebars, and its large yellow-orange beak with a bulging orange throat pouch pointing ahead; behind it, white speed lines show motion, and above, a yellow sun with a pale halo and two white clouds sit in a gradient blue sky.

For comparison, here’s the pelican I got a year ago from Haiku 4.5 (for 0.7583 cents—Haiku 4.5 did not support reasoning levels). It sucked at drawing pelicans:

Described by Haiku 4.5: A whimsical illustration of a bird with a round tan body, pink beak, and orange legs riding a bicycle against a blue sky and green grass background.

And a generous API credit scheme for subscribers

In addition to Haiku 5.5, Anthropic announced today that they are halving the price of cache reads for Sonnet 5.5. They’ve also added API credits to subscription plans:

Second, this week, we’ll roll out a new monthly API credit to all Max and Team subscribers for use on the Claude Platform. Max 5x users will get $100 in credits per month, Max 20x users will get $200, and Team subscribers will receive up to $500, pooled across their users.

Claiming this is pleasantly easy: navigate to Settings -> Billing and select the API organization that should benefit from the credits every month:

Screenshot of a Settings dialog with a close X button at top right and a row of tabs reading "General", "Account", "Privacy", "Billing" (selected), "Usage", "Capabilities", "Memory" and "Design sys" (cut off at the edge). Beside a line-drawn icon of a branching plant with circular buds, the panel reads "Max plan", "20x more usage than Pro" and "Your subscription will auto renew on Nov 2, 2026." with an "Adjust plan" button on the right. Below is a section headed "API credits" with a blue "New" badge, reading "Your Max 20x plan includes $200 USD in API credits each month." followed by grey text "Link an API organization to start receiving credits." with a "Link organization" button on the right.

The API credits exactly match the cost of the subscription itself. This is really generous—it makes it much easier for subscribers to use the API. Anthropic also let you disable auto-reload for the API, with the consequence that “API requests will stop when your balance runs out”—exactly what you want if you’re planning to burn through those API credits without risk of a nasty billing surprise.

Note that the monthly credits do not roll over—use them or lose them.

OpenAI still allow you to use your Codex subscription for personal API use, which works out as a better deal for heavy API users. This new credit scheme goes at least some way to overcoming that difference.