[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"article-claude-opus-4-7-release-workflow-vision-zh":3,"article-related-claude-opus-4-7-release-workflow-vision-zh":28,"series-model-release-97f9c411-2849-4a70-9d1e-3bba0dde23bf":81},{"id":4,"slug":5,"title":6,"content":7,"summary":8,"source":9,"source_url":10,"author":11,"image_url":12,"cover_image":12,"category":13,"language":14,"translated_content":11,"related_article_id":15,"keywords":16,"key_takeaways":11,"views":25,"created_at":26,"published_at":27,"topic_cluster_id":11},"97f9c411-2849-4a70-9d1e-3bba0dde23bf","claude-opus-4-7-release-workflow-vision-zh","Claude Opus 4.7 上線：更會做事了","\u003Cp>\u003Ca href=\"https:\u002F\u002Fwww.anthropic.com\u002F\" target=\"_blank\" rel=\"noopener\">Anthropic\u003C\u002Fa> 推出 \u003Ca href=\"https:\u002F\u002Fwww.anthropic.com\u002Fnews\u002Fclaude-opus-4-7\" target=\"_blank\" rel=\"noopener\">Claude Opus 4.7\u003C\u002Fa>。這次不是只拚聊天順不順，而是把長任務、視覺理解、工作流穩定性拉上來。官方很直白，這版就是要把事情做完。\u003C\u002Fp>\u003Cp>講白了，這種升級對開發者更有感。你拿它改程式、讀截圖、整理報告，差一點點的準確率，就可能少掉一輪返工。對企業來說，這差的不是面子，是工時。\u003C\u002Fp>\u003Cp>但代價也很現實。更高解析度輸入、更長輸出，Token 消耗都會上去。你如果是 API 使用者，帳單會先幫你記得這件事。\u003C\u002Fp>\u003Ch2>這次重點，不是更會聊\u003C\u002Fh2>\u003Cp>Anthropic 把 Opus 4.7 的主軸放在高階軟體工程、長時間任務、嚴格指令遵循。這代表模型不只會回你一句漂亮答案，還要能一路把步驟做完。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1776859436510-gf5y.png\" alt=\"Claude Opus 4.7 上線：更會做事了\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>這種方向很合理。現在很多人已經不缺「會說」的模型，缺的是「能收尾」的模型。尤其是文件整理、跨檔案修改、研究摘要這類工作，半路跑掉真的很煩。\u003C\u002Fp>\u003Cp>如果你有用過舊版 Claude，就知道它有時候前半段很穩，後半段開始飄。Opus 4.7 想解的，就是這種長鏈路任務的掉線問題。\u003C\u002Fp>\u003Cul>\u003Cli>\u003Ca href=\"https:\u002F\u002Fwww.anthropic.com\u002Fnews\u002Fclaude-opus-4-7\" target=\"_blank\" rel=\"noopener\">官方發布頁\u003C\u002Fa>主打長任務與代理式工作流\u003C\u002Fli>\u003Cli>SWE-bench Multilingual：80.5%\u003C\u002Fli>\u003Cli>GraphWalks BFS 1M：58.6%\u003C\u002Fli>\u003Cli>Vending-Bench 2：最終餘額 10,937 美元\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>看圖能力，這次補得很兇\u003C\u002Fh2>\u003Cp>這版另一個很實際的升級，是看圖更細。Anthropic 提到，Opus 4.7 支援長邊最高 2576 像素的圖像輸入，約 375 萬像素。對密集截圖、圖表、流程圖、介面原型圖，這很有用。\u003C\u002Fp>\u003Cp>以前很多模型一碰高解析 UI，就會漏小字、漏按鈕、漏局部結構。這次比較像是把「看得到」變成「看得清」。對 Computer Use 場景，這差很多。\u003C\u002Fp>\u003Cp>你如果做產品設計、QA、前端除錯，這種能力很實用。因為它不只是讀圖，而是要從圖裡抓出操作線索。\u003C\u002Fp>\u003Cblockquote>“The future is already here — it’s just not very evenly distributed.” — William Gibson\u003C\u002Fblockquote>\u003Cp>這句話放在 \u003Ca href=\"\u002Fnews\u002Fai-papers-of-the-week-ml-paper-roundup-zh\">AI\u003C\u002Fa> 很貼切。有人還在拿模型聊天，有人已經拿它讀截圖、找欄位、整理表格。Opus 4.7 讓這條線又往前挪了一點。\u003C\u002Fp>\u003Ch2>和競品比，差距開始變具體\u003C\u002Fh2>\u003Cp>只看單一版本，Opus 4.7 只是比 Opus 4.6 強一截。但把它放進同級比較，差距就很清楚了。\u003Ca href=\"https:\u002F\u002Fartificialanalysis.ai\u002Fevals\u002Fgdpval-aa\" target=\"_blank\" rel=\"noopener\">Artificial Analysis 的 GDPval-AA\u003C\u002Fa> 評估，涵蓋 44 種知識工作職業與 9 大產業，任務來自平均 14 年經驗的資深從業者。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1776859440187-hq79.png\" alt=\"Claude Opus 4.7 上線：更會做事了\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>在這份評估裡，Opus 4.7 的 Elo 是 1753。Opus 4.6 是 1619。\u003Ca href=\"https:\u002F\u002Fopenai.com\u002Findex\u002Fgpt-5-4\u002F\" target=\"_blank\" rel=\"noopener\">GPT-5.4\u003C\u002Fa> 是 1674。\u003Ca href=\"https:\u002F\u002Fwww.google.com\u002Fintl\u002Fen\u002Fai\u002Fgemini\u002F\" target=\"_blank\" rel=\"noopener\">Gemini 3.1 Pro\u003C\u002Fa> 是 1314。這組數字很直接，Opus 4.7 已經把不少只會寫漂亮話的模型甩開。\u003C\u002Fp>\u003Cp>企業文件推理的差距更誇張。\u003Ca href=\"https:\u002F\u002Fwww.databricks.com\u002Fblog\u002Fofficeqa-pro-benchmark\" target=\"_blank\" rel=\"noopener\">Databricks OfficeQA Pro\u003C\u002Fa> 測的是接近 100 年的美國財政部公報，資料有 8.9 萬頁 PDF 和 2600 萬個數字。這種題目很吃耐心，也很吃上下文管理。\u003C\u002Fp>\u003Cul>\u003Cli>GDPval-AA：Opus 4.7 1753，GPT-5.4 1674，Gemini 3.1 Pro 1314\u003C\u002Fli>\u003Cli>OfficeQA Pro：Opus 4.7 80.6%，Opus 4.6 57.1%\u003C\u002Fli>\u003Cli>Structural Biology：Opus 4.7 74.0%，Opus 4.6 30.9%\u003C\u002Fli>\u003Cli>SWE-bench Multimodal：Opus 4.7 34.5%，Opus 4.6 27.1%\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>成本和安全，還是不能跳過\u003C\u002Fh2>\u003Cp>Opus 4.7 不是白送的升級。Anthropic 明講了，高解析度圖像會吃更多 Token，新分詞器也可能讓同樣輸入變成更多 Token。高 Effort 模式下，輸出也會更長。\u003C\u002Fp>\u003Cp>這對個人用戶是額度問題。對團隊和 API 用戶，就是成本問題。你如果一天跑幾百次工作流，差一點點 Token，月底就會很有感。\u003C\u002Fp>\u003Cp>\u003Ca href=\"\u002Fnews\u002Fsafe-continual-rl-changing-environments-zh\">安全\u003C\u002Fa>面也不能忽略。Anthropic 在發布前提過 \u003Ca href=\"https:\u002F\u002Fwww.anthropic.com\u002Fnews\u002Fproject-glasswing\" target=\"_blank\" rel=\"noopener\">Project Glasswing\u003C\u002Fa>，談前沿模型在網安上的風險與收益。Opus 4.7 也帶有自動偵測與攔截高風險網安請求的護欄。\u003C\u002Fp>\u003Cp>這代表它不是只往能力衝，也在收\u003Ca href=\"\u002Fnews\u002Fedge-of-stability-generalization-zh\">邊界\u003C\u002Fa>。說真的，這比空喊口號實際多了。\u003C\u002Fp>\u003Ch2>這版會先影響誰？\u003C\u002Fh2>\u003Cp>先有感的，大概不是只拿來聊天的人，而是每天都在處理程式、表格、截圖、文件的人。因為它的價值不在文采，而在少跑偏、少返工、少人工盯著。\u003C\u002Fp>\u003Cp>如果你的流程本來就會讓模型先出草稿，再人工校對，Opus 4.7 會比較像一個能穩定接手中段工作的助手。它不一定每次都驚艷，但它可能更少搞砸。\u003C\u002Fp>\u003Cp>我覺得接下來真正值得看的是企業採用率。問題已經不是模型能不能寫，而是它能不能在你的流程裡，乖乖把事情寫完。\u003C\u002Fp>\u003Cp>如果你是開發者，現在就該測兩件事：長上下文穩不穩，Token 成本高不高。這兩個答案，會直接決定你要不要把它放進正式流程。\u003C\u002Fp>","Anthropic 推出 Claude Opus 4.7，強化長任務、視覺理解與程式工作流，但 Token 消耗也更高。","zhuanlan.zhihu.com","https:\u002F\u002Fzhuanlan.zhihu.com\u002Fp\u002F2028396026466247335",null,"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1776859436510-gf5y.png","model-release","zh","2ab61916-02e3-47f5-8131-9d69cb770f03",[17,18,19,20,21,22,23,24],"Claude Opus 4.7","Anthropic","LLM","Token","API","視覺理解","長任務","AI 工作流",6,"2026-04-22T12:03:36.654584+00:00","2026-04-22T12:03:36.457+00:00",{"tags":29,"relatedLang":40,"relatedPosts":44},[30,32,34,36,38],{"name":18,"slug":31},"anthropic",{"name":24,"slug":33},"ai-工作流",{"name":19,"slug":35},"llm",{"name":37,"slug":37},"token",{"name":21,"slug":39},"api",{"id":15,"slug":41,"title":42,"language":43},"claude-opus-4-7-release-workflow-vision-en","Claude Opus 4.7 发布：更会干活了","en",[45,51,57,63,69,75],{"id":46,"slug":47,"title":48,"cover_image":49,"image_url":49,"created_at":50,"category":13},"cfb68e08-fe4e-49f6-b449-e566faf56311","kimi-2-7-price-coding-benchmark-zh","Kimi 2.7 讓價格成為真正的寫碼基準","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782746270225-tcs9.png","2026-06-29T15:17:24.321277+00:00",{"id":52,"slug":53,"title":54,"cover_image":55,"image_url":55,"created_at":56,"category":13},"ca1e6960-10e7-4fa7-949f-c5991c99fc7e","kimi-k26-open-source-coding-agentic-ai-benchmarks-zh","Kimi K2.6 登頂程式與代理式 AI 基準","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782739078329-qvne.png","2026-06-29T13:17:26.530857+00:00",{"id":58,"slug":59,"title":60,"cover_image":61,"image_url":61,"created_at":62,"category":13},"edf8e66b-c717-4cc1-b15a-96839bb7bbcf","llama-legends-380-season-3-heroes-raids-zh","Llama Legends 3.8.0 推出 Season 3 英雄與突襲","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782711179415-qurv.png","2026-06-29T05:32:32.733919+00:00",{"id":64,"slug":65,"title":66,"cover_image":67,"image_url":67,"created_at":68,"category":13},"88d353ca-468b-4774-922d-ef0cbc2edd68","omlx-045-dev1-glm52-minimax-m3-speedups-zh","oMLX 0.4.5.dev1 讓長上下文更快","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782709372375-25nm.png","2026-06-29T05:02:28.341041+00:00",{"id":70,"slug":71,"title":72,"cover_image":73,"image_url":73,"created_at":74,"category":13},"e6ae84b6-4e55-4ab2-a1cf-4a08e23cbc77","grok-45-private-beta-tesla-spacex-zh","Grok 4.5 先進 Tesla 和 SpaceX 內測","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782687769532-te5b.png","2026-06-28T23:02:22.915901+00:00",{"id":76,"slug":77,"title":78,"cover_image":79,"image_url":79,"created_at":80,"category":13},"186b266a-5b45-4bd4-85a4-5fa62fcc50dc","google-openrl-llm-fine-tuning-kubernetes-zh","Google OpenRL 把 RL 細調搬上 Kubernetes","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782572576166-gzxw.png","2026-06-27T15:02:27.036919+00:00",[82,87,92,97,102,107,112,117,122,127],{"id":83,"slug":84,"title":85,"created_at":86},"58b64033-7eb6-49b9-9aab-01cf8ae1b2f2","nvidia-rubin-six-chips-one-ai-supercomputer-zh","NVIDIA Rubin 把六顆晶片塞進 AI 機櫃","2026-03-26T07:18:45.861277+00:00",{"id":88,"slug":89,"title":90,"created_at":91},"0dcc2c61-c2a6-480d-adb8-dd225fc68914","march-2026-ai-model-news-what-mattered-zh","2026 年 3 月 AI 模型新聞重點","2026-03-26T07:32:08.386348+00:00",{"id":93,"slug":94,"title":95,"created_at":96},"214ab08b-5ce5-4b5c-8b72-47619d8675dd","why-small-models-are-winning-on-device-ai-zh","小模型為何吃下裝置端 AI","2026-03-26T07:36:30.488966+00:00",{"id":98,"slug":99,"title":100,"created_at":101},"785624b2-0355-4b82-adc3-de5e45eecd88","midjourney-v8-faster-images-higher-costs-zh","Midjourney V8 變快了，也變貴了","2026-03-26T07:52:03.562971+00:00",{"id":103,"slug":104,"title":105,"created_at":106},"cda76b92-d209-4134-86c1-a60f5bc7b128","xiaomi-mimo-trio-agents-robots-voice-zh","小米 MiMo 三模型瞄準代理、機器人與語音","2026-03-28T03:05:08.779489+00:00",{"id":108,"slug":109,"title":110,"created_at":111},"9e1044b4-946d-47fe-9e2a-c2ee032e1164","xiaomi-mimo-v2-pro-1t-moe-agents-zh","小米 MiMo-V2-Pro 登場：1T MoE 模型","2026-03-28T03:06:19.002353+00:00",{"id":113,"slug":114,"title":115,"created_at":116},"c4b6186f-bd84-4598-997e-c6e31d543c0d","cursor-composer-2-agentic-coding-model-zh","Cursor Composer 2 走向代理式寫碼","2026-03-28T03:13:06.422716+00:00",{"id":118,"slug":119,"title":120,"created_at":121},"e112e76f-ec3b-408f-810e-e93ae21a888a","apple-siri-gemini-distilled-models-zh","Apple Siri 牽手 Gemini 的真相","2026-03-29T04:52:57.886544+00:00",{"id":123,"slug":124,"title":125,"created_at":126},"c679b51f-194a-463b-87fc-7695256ff752","mimo-v2-pro-vs-omni-vs-flash-2026-zh","MiMo V2 Pro、Omni、Flash 怎麼選","2026-04-02T01:18:43.576128+00:00",{"id":128,"slug":129,"title":130,"created_at":131},"3b988fd7-6749-4f01-ba25-c0ad7486dc31","z-ai-glm-5v-turbo-design2code-claude-zh","GLM-5V-Turbo 在 Design2Code 贏了…","2026-04-02T04:03:36.31741+00:00"]