Anthropic 与 OpenAI 新模型作出同样承诺：以大幅更低成本提供小幅提升

英文标题：New Anthropic, OpenAI models make same promise: A little more for a lot less money

Anthropic 表示，Opus 5.5 在默认设置下处理典型工作负载时，相比 Opus 5 的节省幅度可接近 40%，原因不仅是 token 单价下降，也包括完成任务所需 token 更少。由于 Opus 5.5 在网络安全和生物学等高风险领域能力突出，其请求仍可能被自动且透明地路由至旧模型。OpenAI 则发布了 GPT-6 Sol 与 Luna，定位为 GPT-6 Astra 之外更高效、更低价的选择。两款模型采用与 Astra 相似的训练方法，部分基准表现较前代提升数个百分点，但使用成本降至原来的一半。

打开原文相关 2 · 机会 0

中文

Anthropic 进一步表示，在默认设置下，对于典型工作负载，Opus 5.5 相比 Opus 5 的成本节省幅度接近 40%。原因不仅在于 token 成本下降，也在于 Opus 5.5 完成任务时使用的 token 更少。

据称，Opus 5.5 在网络安全和生物学等“高风险领域”具备显著能力，因此此前适用于 Fable 5.1 的同类保护措施也将适用于该模型——如果你的请求被标记为涉及受保护领域，系统可能会自动且透明地将其路由至旧版模型。

GPT-6 Sol 与 Luna：一次温和、以成本为重点的升级

OpenAI 发布 GPT-6 Sol 和 Luna，是一次渐进式的进步。不久前，该公司刚为其 GPT-5.6 模型家族引入 Sol、Terra、Luna 的命名体系。而就在最近——本月内——该公司又推出了 GPT-6 Astra，这是其迄今最先进、能力也最强大的模型。

这些名称的含义或许并不直观，以下是简要说明。Astra 是性能最激进、价格也最高的模型，面向高强度编程、研究等任务。Sol 则是能力出色但更高效、更实惠的替代方案，定位为处理此类任务的日常主力模型。Terra 是均衡的通用模型。而 Luna 则是快速、低价的选择。

可以大致将 Astra 对应 Anthropic 的 Fable、Sol 对应 Opus、Terra 对应 Sonnet，以及 Luna 对应 Haiku，但这并非完全准确的映射，尤其对于高端模型而言。

OpenAI 表示，GPT-6 Sol 和 Luna 采用了与 GPT-6 Astra 类似的训练方法。根据不同基准测试，它们在某些任务上的能力有时比前代高出几个百分点，但使用成本仅为前代的一半。

English / 原文

Anthropic further claims that the savings are closer to 40 percent compared to Opus 5 for typical workloads at default settings, because in addition to the cost of tokens going down, Opus 5.5 also uses fewer tokens when completing tasks.

Opus 5.5 is said to be notably capable in “high risk areas” like cybersecurity and biology, so the same protections that applied to Fable 5.1 will also apply here—your requests might be automatically and transparently routed to an older model if they get flagged as treading into protected territory.

GPT-6 Sol and Luna: A modest, cost-focused upgrade

The release of OpenAI’s GPT-6 Sol and Luna is an iterative step forward. Not long ago, the company introduced its Sol, Terra, Luna naming convention for its GPT-5.6 family of models. And more recently (just this month), it introduced GPT-6 Astra, which has been both its most advanced and most powerful model.

It may not be obvious what those names mean, so here’s the quick rundown. Astra is the most aggressively powerful (and pricey) model, meant for heavy-duty coding, research, and so on. Sol is a capable but more efficient and affordable alternative—meant to be the sort of daily driver for tasks like that. Terra is the balanced, general-use model. And Luna is the fast, cheap option.

You could reasonably and roughly position Astra against Anthropic’s Fable, Sol against Opus, Terra against Sonnet, and Luna against Haiku, but it would be an imperfect mapping, especially for the higher-end models.

OpenAI says GPT-6 Sol and Luna were trained with similar methods to those used to train GPT-6 Astra. Depending on the benchmark, they’re sometimes a few percentage points more capable than their predecessors at certain tasks, but they cost half as much to use.