New Anthropic, OpenAI models make same promise: A little more for a lot less money
https://arstechnica.com/ai/2026/09/new-anthropic-openai-models-make-same-promise-a-little-more-for-a-lot-less-money/📌 Opus 5.5 對決 GPT-6 Sol／Luna：這場軍備競賽拚的不是能力，是價格
TL;DR：Anthropic Opus 5.5 與 OpenAI Sol／Luna 同天釋出，兩家不約而同選擇用降價而非突破性能力來搶企業市場。
同一週，Anthropic 與 OpenAI 各自端出新模型，卻講著幾乎同一個故事：能力小幅提升，價格大幅下砍。這不是巧合，而是一場企業市場的成本戰。
🤔 為什麼兩大廠同時選擇「降價優先」
Ars Technica 指出，這批新模型的重點不在突破性能力，而在效率。隨著 OpenAI 與 Anthropic 都把企業客戶視為核心戰場，他們正與開放權重模型（open-weight models）正面競爭；不少企業已經開始用 model router 減少對昂貴前沿模型的依賴，改用更便宜的替代方案。兩家公司都主張，這些新版本在前沿能力上有小幅推進，但真正的重點是把成本大幅壓低。
🧩 Opus 5.5：一次「追趕」式的降價與提速
Opus 5.5 是 Anthropic 面向大眾市場的主力工作模型，用於程式開發與複雜知識工作。在此之前，OpenAI 本月稍早發布的 GPT-6 Astra，在部分 benchmark 與使用者評價上已略勝 Opus 5 一籌（就定價與能力而言，Astra 同時與 Opus 與 Fable 競爭）。Anthropic 與合作夥伴的 benchmark 顯示，Opus 5.5 在程式開發與知識工作上，某些情況下已優於 GPT-6 Astra，但幅度不大。
📊 真正的故事是成本
根據 Anthropic 官方公告，Opus 5.5 的輸入與輸出價格分別為每百萬 token $4 與 $20，較 Opus 5 便宜 20%。佔 agentic 與程式工作成本大宗的快取讀取（cache reads）價格為每百萬 token $0.20，較 Opus 5 便宜 60%。此外，Opus 5.5 的輸出速度比 Opus 5 快超過 30%。
💡 同一場戰爭的兩條戰線
OpenAI 這邊則是針對中階與輕量模型出手：GPT-6 Sol 與 Luna 定位為中階／小型模型，主打效率與速度。兩家公司的動作合起來看，勾勒出同一個趨勢，前沿模型的「能力差距」正在縮小、被視為理所當然，接下來比拚的是誰能用更低的每 token 成本，把接近前沿的能力交到企業手上。
🎯 實務啟示
如果你的團隊正在評估要不要導入 model router 或混合模型策略，這輪降價值得重新計算 TCO（總持有成本）。尤其是快取讀取價格的下降，對大量重複 context 的 agentic 應用影響最直接，值得優先納入成本模型重新試算。
🔗 來源
- 標題：New Anthropic, OpenAI models make same promise: A little more for a lot less money
- 作者／機構：Samuel Axon／Ars Technica
- 連結：https://arstechnica.com/ai/2026/09/new-anthropic-openai-models-make-same-promise-a-little-more-for-a-lot-less-money/
#Anthropic #OpenAI #OpusAI #GPT6 #AIΡricing #EnterpriseAI #LLMCompetition #ModelRouting #AIInfrastructure #MachineLearning
原始資料 Ars Technica AI · 收集於 2026-09-23
摘要原文
OpenAI and Anthropic both recently released new models aimed at lowering costs. Anthropic announced Opus 5.5, the latest version of its main mass-market workhorse model, used for tasks like coding and other complex knowledge work. And OpenAI announced GPT-6 Sol and Luna, the latest versions of its middle-of-the-road or smaller models focused on efficiency and speed. These new releases are not about groundbreaking new capabilities. Rather, they’re about efficiency. As both OpenAI and Anthropic target enterprise customers, they’re racing to compete with open-weight models as organizations have explored changing their practices and using model routers to use these pricey frontier models less in favor of cheaper alternatives. Anthropic and OpenAI argue that these new releases push the envelope at the frontier (albeit mostly in modest ways), while bringing costs substantially down. Opus 5.5: A more affordable, more competitive flagship Opus 5.5 is at the higher end of the models announced today, but the wider context here is that Anthropic is playing a bit of catch-up in its race with OpenAI. OpenAI earlier this month released GPT-6 Astra, which has sometimes been modestly beating Opus 5 in benchmarks and user sentiment. (By price and capability, Astra is competing with both Opus and Fable.) Benchmarks by Anthropic and its partners now show Opus 5.5 performing better at coding and knowledge work than GPT-6 Astra in some cases, albeit modestly. The real story here is cost. From Anthropic’s announcement : “Input and output tokens are $4 and $20 per million, 20% less than Opus 5. Cache reads (which make up the majority of agentic and coding work costs) are $0.20 per million tokens, 60% less than Opus 5. Opus 5.5 also generates output more than 30% faster than Opus 5.”
由 claude-code/sonnet 自動生成