[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"article-5-shifts-in-llms-from-the-last-six-months-zh":3,"article-related-5-shifts-in-llms-from-the-last-six-months-zh":38,"series-industry-be0785a5-7976-4735-8f46-6abd84dac9af":91},{"id":4,"title":5,"content":6,"summary":7,"source":8,"source_url":9,"author":10,"image_url":11,"keywords":12,"language":21,"translated_content":10,"views":22,"is_premium":23,"created_at":24,"updated_at":24,"cover_image":11,"published_at":25,"rewrite_status":26,"rewrite_error":10,"rewritten_from_id":27,"slug":28,"category":29,"related_article_id":30,"status":31,"google_indexed_at":10,"x_posted_at":10,"tweet_text":10,"title_rewritten_at":10,"title_original":10,"key_takeaways":32,"topic_cluster_id":36,"embedding":37,"is_canonical_seed":23},"be0785a5-7976-4735-8f46-6abd84dac9af","5 個 LLM 的半年轉變","\u003Cp data-speakable=\"summary\">這篇整理六個月內 LLM 的五個關鍵轉變，幫你判斷該用雲端前沿模型、開放模型，還是本地工作流。\u003C\u002Fp>\u003Cp>在短短 6 個月內，LLM 的使用方式明顯改變。讀完這 5 項，你可以更快決定：該把預算放在 coding agent、開放模型，還是本地部署。\u003C\u002Fp>\u003Ctable>\u003Cthead>\u003Ctr>\u003Cth>項目\u003C\u002Fth>\u003Cth>代表規格\u003C\u002Fth>\u003Cth>實際意義\u003C\u002Fth>\u003C\u002Ftr>\u003C\u002Fthead>\u003Ctbody>\u003Ctr>\u003Ctd>Coding agents\u003C\u002Ftd>\u003Ctd>從「常常要修」變成「大多可直接用」\u003C\u002Ftd>\u003Ctd>適合日常開發、重構、測試\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>最佳模型更替\u003C\u002Ftd>\u003Ctd>數月內多次換位\u003C\u002Ftd>\u003Ctd>不能只看靜態排行榜\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>本地／開放模型\u003C\u002Ftd>\u003Ctd>20.9GB 到 1.5TB\u003C\u002Ftd>\u003Ctd>可在筆電到大型伺服器間選擇\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>個人 AI 助手\u003C\u002Ftd>\u003Ctd>Mac mini 常見\u003C\u002Ftd>\u003Ctd>適合長期、持續性任務\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Benchmark demos\u003C\u002Ftd>\u003Ctd>像 pelican test 這類測試\u003C\u002Ftd>\u003Ctd>更能看出多模態與工具鏈能力\u003C\u002Ftd>\u003C\u002Ftr>\u003C\u002Ftbody>\u003C\u002Ftable>\u003Ch2>1. Coding agents 夠日常使用了\u003C\u002Fh2>\u003Cp>最大的變化，不是某一個模型單獨升級，而是 \u003Ca href=\"\u002Ftag\u002Fagentic-coding\">agentic coding\u003C\u002Fa> 的整體品質跳了一級。像 \u003Ca href=\"https:\u002F\u002Fopenai.com\u002Fcodex\u002F\" rel=\"nofollow\">Codex\u003C\u002Fa> 和 \u003Ca href=\"https:\u002F\u002Fclaude.ai\u002Fcode\" rel=\"nofollow\">Claude Code\u003C\u002Fa> 這類工具，開始能在可驗證的回饋下持續優化，結果就是輸出不再像展示品，而是真的能幫忙做事。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1779167646556-uxgd.png\" alt=\"5 個 LLM 的半年轉變\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>這代表工作流也變了。以前你花大部分時間在修錯，現在可以把完整任務交出去，再回來檢查結果。\u003Ca href=\"\u002Fnews\u002F5-indiana-fever-updates-zh\">重點\u003C\u002Fa>不是完美，而是「大多可用」已經足夠進入日常開發。\u003C\u002Fp>\u003Cul>\u003Cli>之前：常常要手動修補\u003C\u002Fli>\u003Cli>之後：大多能直接接手\u003C\u002Fli>\u003Cli>適合：寫程式、重構、補測試、小型功能\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>2. 「最佳模型」換位變得很頻繁\u003C\u002Fh2>\u003Cp>另一個明顯現象，是領先位置在幾個大廠之間快速輪替。短短幾個月，冠軍從 \u003Ca href=\"\u002Ftag\u002Fclaude\">Claude\u003C\u002Fa> Sonnet 4.5、到 GPT-5.1、到 \u003Ca href=\"\u002Ftag\u002Fgemini\">Gemini\u003C\u002Fa> 3、再到 GPT-5.1 Codex Max，\u003Ca href=\"\u002Fnews\u002Fbree-hall-returns-indiana-fever-player-development-zh\">最後\u003C\u002Fa>又回到 \u003Ca href=\"\u002Ftag\u002Fanthropic\">Anthropic\u003C\u002Fa> 的 Claude Opus 4.5。\u003C\u002Fp>\u003Cp>這種變動說明競爭已經非常接近。對實際使用者來說，最好的做法不是相信固定排名，而是拿自己的任務去測。擅長寫 code 的模型，不一定最會做長流程規劃，也不一定最適合影像或工具調用。\u003C\u002Fp>\u003Cul>\u003Cli>Claude Sonnet 4.5：前段時間的領先者\u003C\u002Fli>\u003Cli>GPT-5.1 與 GPT-5.1 Codex Max：中段強勢競爭者\u003C\u002Fli>\u003Cli>Gemini 3：在特定測試上表現突出\u003C\u002Fli>\u003Cli>Claude Opus 4.5：後段重新奪回優勢\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>3. 開放模型和本地模型進步很大\u003C\u002Fh2>\u003Cp>開放模型這邊也跑得很快。像 \u003Ca href=\"https:\u002F\u002Fai.google.dev\u002Fgemma\" rel=\"nofollow\">Gemma 4\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fglm.ai\u002F\" rel=\"nofollow\">GLM-5.1\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fqwenlm.github.io\u002F\" rel=\"nofollow\">Qwen3.6-35B-A3B\u003C\u002Fa> 這些模型，證明 local 或 self-hosted 已經不只是退而求其次，而是可行選項。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1779167648208-i8c2.png\" alt=\"5 個 LLM 的半年轉變\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>真正改變的是「能力對尺寸」的比值。20.9GB 的模型在筆電上就能跑出超乎預期的結果，而 1.5TB 的大模型則能在足夠硬體下展現很強表現。現在的問題不再是能不能本地跑，而是你要選哪種成本與能力的平衡。\u003C\u002Fp>\u003Cul>\u003Cli>Gemma 4：作者認為最強的美系開放模型之一\u003C\u002Fli>\u003Cli>GLM-5.1：體積大、吃硬體，但能力強\u003C\u002Fli>\u003Cli>Qwen3.6-35B-A3B：相對筆電友善\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>4. 個人 AI 助手開始成形\u003C\u002Fh2>\u003Cp>原本不起眼的 Warelay repo，後來變成 \u003Ca href=\"https:\u002F\u002Fopenclaw.ai\u002F\" rel=\"nofollow\">OpenClaw\u003C\u002Fa>，到 2 月時已經吸引大量關注。更重要的是，這類工具開始有了更通用的稱呼：Claws，也就是以 agentic pattern 為核心的個人 AI 助手。\u003C\u002Fp>\u003Cp>這件事重要在於，它把互動模式從一次性的聊天，推進到持續存在的助手。有人甚至會特地買 Mac mini 來跑這些系統。概念很簡單：準備一台專門的機器，讓助手長期處理任務，不要占用主力電腦。\u003C\u002Fp>\u003Cul>\u003Cli>Warelay：最初的 repo 名稱\u003C\u002Fli>\u003Cli>OpenClaw：後來定名並被廣泛注意\u003C\u002Fli>\u003Cli>Claws：逐漸形成的類別名稱\u003C\u002Fli>\u003Cli>常見配置：一台 Mac mini 當助手主機\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>5. Benchmark demo 變得更怪，也更有參考價值\u003C\u002Fh2>\u003Cp>Simon Willison 的 pelican-riding-a-bicycle \u003Ca href=\"\u002Fnews\u002F5-llm-benchmarks-for-business-buyers-2026-zh\">測試\u003C\u002Fa>，會一直被拿來比較模型，因為它荒謬得剛剛好。它難畫、容易辨識，而且不太可能被各家模型專門優化，所以很適合用來看多模態表現。\u003C\u002Fp>\u003Cp>同一時期也出現很多有趣但有意義的 demo，例如用 \u003Ca href=\"https:\u002F\u002Fpyodide.org\u002F\" rel=\"nofollow\">Pyodide\u003C\u002Fa> 在瀏覽器裡跑 WebAssembly，再用 Python 包住 JavaScript。這些例子看起來像玩具，但其實證明了工具鏈已經成熟到足以支持奇怪的實驗。\u003C\u002Fp>\u003Cpre>\u003Ccode>browser → JavaScript → WebAssembly → Pyodide → Python → micro-javascript\u003C\u002Fcode>\u003C\u002Fpre>\u003Cul>\u003Cli>Pelican test：快速檢查模型品質\u003C\u002Fli>\u003Cli>Micro-javascript：小型但很有說服力的實驗\u003C\u002Fli>\u003Cli>重點：工具夠成熟，才會有這些怪但有用的 demo\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>怎麼挑\u003C\u002Fh2>\u003Cp>如果你最在意寫程式效率，先試 \u003Ca href=\"\u002Ftag\u002Fopenai\">OpenAI\u003C\u002Fa> 和 Anthropic 的 agentic 工具，再用自己的 repo 測。若你重視隱私、成本或離線使用，就看 Google、GLM 和 Qwen 的開放模型。若你在做產品，核心判斷已經不是「能不能做」，而是「哪個模型最適合這個任務、這台機器和這個預算」。\u003C\u002Fp>\u003Cp>對多數人來說，最實際的組合是保留一個前沿模型，再配一個本地模型。前者處理難題，後者處理日常工作，會最穩。\u003C\u002Fp>","5 個轉變說明 LLM 為何在 6 個月內快速改變：更強的 coding agents、開放模型與本地工作流。","simonwillison.net","https:\u002F\u002Fsimonwillison.net\u002F2026\u002FMay\u002F19\u002F5-minute-llms\u002F",null,"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1779167646556-uxgd.png",[13,14,15,16,17,18,19,20],"LLM","coding agents","open models","local workflows","Claude","GPT","Gemini","OpenClaw","zh",2,false,"2026-05-19T05:13:34.442673+00:00","2026-05-19T05:13:34.431+00:00","done","7d533544-e448-4f47-acf0-72e0ac21c2ad","5-shifts-in-llms-from-the-last-six-months-zh","industry","bc80ab6d-7eba-4d56-9e4a-d58aa61328cb","published",[33,34,35],"LLM 在半年內的最大變化，是 coding agents 從展示品變成可日常使用。","模型排行榜變動很快，實際任務測試比固定排名更重要。","開放模型與本地部署的可用性提升，讓雲端與離線工作流都更成熟。","caa87b65-9bbc-46fe-bba8-4f4158dd2d8b","[-0.023804283,0.015670937,0.0031465704,-0.05863659,0.000107499574,0.00372833,-0.027072728,0.023346318,0.0056801327,-0.0031506969,0.005637778,-0.014330226,0.019630147,-0.022717725,0.13451524,0.048291966,-0.0022833168,0.029433789,-0.002121879,0.01000842,0.009460498,-0.000394414,0.011520271,-0.013110213,-0.014012565,0.009784582,0.017515667,0.0039119665,0.04578385,0.006115479,-0.0028564383,0.019805973,0.00017468895,0.013225934,-0.0024900858,0.010501677,0.022172986,-0.0052270545,0.022871355,0.0017302458,-0.01068783,-0.015618173,0.011233779,-0.025555879,-0.02472139,0.00761643,-0.008545043,-0.03346948,-0.03622226,0.018055037,-0.031574406,0.0144642545,0.010729533,-0.14893304,-0.0066143465,0.012364936,-0.010737088,0.016362023,0.014949989,0.0106458925,-0.0059311953,0.016971182,-0.031032493,0.0064428183,-0.005088524,-0.043828618,0.032894634,-0.007331878,-0.012627402,-0.0047067404,-0.012716459,0.02219621,0.01175148,-0.037935555,0.0022372599,-0.036200956,0.0043310076,0.029712643,0.010484576,0.020590981,0.013725608,-0.0058156587,-0.01849999,0.022690464,0.01218502,-0.007014198,-0.0069248006,-0.023479832,-0.00596335,0.0083175935,0.0016565847,-0.004579086,0.008245753,0.014692666,-0.009371281,-0.008816643,-0.008335872,-0.009453138,0.015471728,-0.027788442,0.00416691,-0.0025865445,0.012398471,0.026545169,0.0040364414,-0.005001714,-0.0046164175,0.028401239,0.008079854,0.009550368,-0.02094802,-0.007079612,-0.014969351,-1.9268873e-05,-0.005137,-0.13432641,0.010738775,0.004107692,-0.0041512456,-0.022563864,0.0034352213,0.04047831,-0.002802739,0.018976785,-0.02569992,-0.012944584,0.025438454,-0.014578525,-0.013792093,0.014442146,-0.06252413,0.0021555428,-0.01160013,-0.009424438,-0.0091672,0.010580677,-0.016409121,-0.0038885474,-0.023921734,-0.024279041,-0.005765022,0.038816936,-0.0049987044,-0.0068901815,-0.020955617,0.005551861,-0.03918727,-0.01118646,-0.006647423,0.00936387,0.020570967,-0.0031416009,-0.015307372,-0.009219819,0.022596385,-0.013474495,0.009341248,0.021894118,0.019986527,0.02359299,-0.0075216307,-0.0009220984,0.0216199,0.009310195,-0.0034392856,0.021025447,-0.040729955,0.025694098,0.009536545,0.018469976,-0.008441561,-0.008062644,0.00087311596,0.0026077486,0.010261833,-0.0031501574,-0.022695309,-0.006521946,0.026569339,-0.0026763245,0.020966033,0.0062942253,0.009601356,0.01672779,0.0031940797,-0.021562014,-0.0002291298,0.005944485,-0.0012082127,0.015662875,-0.012230659,0.0152907735,0.024404531,-0.0011144374,-0.010260863,0.0031160025,0.0070844507,0.008637828,-0.003533181,0.01846099,0.025389452,-0.0056111477,0.018169759,-0.05116943,0.02264776,-0.0028868043,-0.01066101,0.010014965,0.004895636,-0.0011065525,-0.013682225,0.040219076,0.026372923,-0.034314547,-0.006169063,-0.0029926,-0.0055309506,-0.017842319,0.013011828,-0.008755244,-0.0046002,-0.0104516735,0.014038787,0.0060758065,-0.0233832,-0.0178149,0.013098697,-0.0035780214,-0.027026214,0.02804902,0.01858337,0.016976107,-0.0067200656,0.017392496,0.013025221,0.021086978,-0.0014713058,0.012671582,-0.004102986,0.035652503,-0.01651669,0.013608892,0.006115831,0.016308196,0.031558726,-0.036574148,0.032596286,0.0004568278,-0.044985864,0.024223126,-0.0041103475,0.0018202892,-0.00483996,0.000562179,-0.0017590167,-0.017913604,0.009348862,0.03505397,-0.015964199,0.008189078,-0.0001564456,0.008228804,-0.0027355722,-0.0077678147,0.005227784,0.0038489278,0.013894787,0.0052410774,-0.022863228,-0.027273066,0.013934476,0.03171216,0.003055742,-0.005825735,0.008209591,-0.019404188,-0.035894763,0.053058244,0.02481125,-0.0011827265,0.01619196,0.004210656,0.0023491434,0.024790969,-0.018524418,-0.014631074,-0.011644342,-0.005386981,-0.007697161,0.0012661497,0.01828586,0.015242968,-0.008159688,-0.01649154,-0.01979092,-0.021256093,0.002030004,0.0064460454,0.0067097642,-0.0029637315,-0.008135803,-0.009042299,0.003357402,0.04198126,-0.032124862,-0.0033044526,0.012729906,0.019300887,-0.022572825,-0.007942163,0.013179046,-0.01683576,0.0031912725,-0.03794993,0.0023730674,-0.013902428,0.026225705,0.0009709129,-0.022903414,0.01939351,-0.01820998,-0.029053148,0.00042141127,-0.0058660875,-0.007960249,0.005423235,-0.0001647109,0.0040921574,0.0043309554,-0.008463965,-0.0118795615,-0.0039984384,0.030178217,0.009729747,-0.0026211003,-0.002757746,-0.009731957,0.002392084,-0.017542588,0.021027474,-0.0043342095,-0.0044667926,-0.023006093,-0.0025763272,-0.0015279079,0.038700376,-0.007850879,0.0043826653,0.004605736,-0.030816548,0.030245127,-0.019672772,-0.0065092677,-0.022832653,-0.02744216,0.024470873,0.0006336287,-0.017158875,0.021113263,-0.010219299,0.009821595,-0.01363803,-0.006864294,-0.01855413,0.020509707,-0.03333287,0.008644064,0.02315815,-0.007663026,-0.0083116675,0.015600662,0.0008567651,-0.0012504212,0.0016984238,-0.020558987,0.015717631,-0.011991245,0.009575764,-0.012599403,-0.004703326,-0.0012945122,0.03315519,-0.019425116,-0.017693056,-0.02711273,0.00017437914,0.026816988,-0.0017286858,-0.0057262722,-0.013816909,-0.01230402,0.01404144,0.010391765,-0.00037209925,-0.008968781,-0.007053656,-0.036173794,0.034990538,-0.0019764353,-0.01821969,0.02271476,0.023138018,-0.015412149,0.013826576,-0.013300969,0.029411245,0.003986559,0.007836641,-0.0038192915,-0.003924701,-0.0031680053,0.023086093,0.004152119,-0.011659961,0.038036883,0.00659147,-0.028474184,0.021990022,0.007819037,-0.0026654229,-0.014734031,0.010093348,-0.011025212,0.014473449,-0.0006484758,-0.014502657,-0.0045999694,-0.0041616205,0.015032063,-0.0070143035,0.008372862,-0.0067786677,-0.023183811,-0.001728562,-0.013092951,-0.0077884216,0.0013006433,-0.017873889,-0.015881801,-0.026956022,0.019632328,0.013966113,0.010476327,0.003272919,0.01562458,0.005418666,-0.0016415338,-0.014334509,-0.059429217,0.021865666,0.026325328,-0.0050331405,0.02699075,-0.01380408,0.0024060477,0.0022948466,-0.014243741,0.029985912,0.011410288,-0.016070276,-0.0036799249,-0.008330314,0.025235219,3.8482394e-05,0.02162778,-0.0036766655,0.0050738947,0.0070272647,-0.0011241648,0.02836408,-0.007707465,-0.0012281705,0.02703171,0.0131341135,-0.006621567,0.0063296547,-0.003383217,-0.022828685,-0.014109129,-0.016866256,-0.0125640975,-0.0252574,0.022189017,-0.020201173,0.007253135,0.00023287364,-0.017715514,-0.0068794675,-0.027395982,0.0040017357,0.032651987,0.026084665,0.017454475,-0.008147988,0.0051552444,0.0034713594,-0.014516754,-0.021215089,-0.006236138,-0.0014129975,-0.022683578,-0.016157528,-0.0024980842,-0.028764091,-0.005452155,-0.013297675,0.0054086447,-0.022596898,-0.007696897,-0.0124371,0.006476954,0.0017163486,0.0054103346,0.021295141,-0.008691933,-0.028499022,0.0073762825,0.007554335,-0.012877432,-0.009262851,-0.0021980244,-0.023500832,-0.0245247,0.0104767345,-0.015561675,-0.0074683516,0.04903222,0.014260215,0.025754783,0.015696019,-0.008442685,0.034530204,-0.0036417875,-0.017536161,0.013687584,0.00023357963,0.029212011,0.04062832,-0.0019275307,0.015839836,-0.031426985,-0.0076299445,-0.011162668,-0.015739396,0.02285165,-0.110958286,-0.0012050231,-0.00057297276,-0.013505984,0.013398449,0.008574014,-0.0046365266,-0.019626686,-0.010855159,0.012167303,0.012291704,-0.0060477913,0.00022694957,0.0013837838,0.0009082546,0.00030088332,-0.015736766,-0.018831454,0.038460586,-0.015149796,0.032535322,0.007251825,0.029794777,0.0056568864,0.018932788,-0.0029045567,0.012836806,0.013710317,-0.0037420986,0.00041610783,-0.030141003,0.006151154,0.0063669756,-0.0027053126,0.007235192,-0.008471924,0.007759687,-0.0034127892,0.026710337,0.002225668,-0.0002462161,-0.007951483,0.012466683,-0.027514324,-0.016749918,-0.0063916054,0.010873717,0.0028143488,0.015508428,0.018652968,-0.03719167,-0.022552049,0.004221269,-0.017950676,-0.007665425,0.004698861,-0.01861472,-0.0022466949,-0.017498882,0.026098348,0.0015345064,-0.0012592964,0.004952547,0.0006392227,-0.014356026,0.0016909023,-0.023988822,0.015141295,-0.018817063,0.0012398928,-0.012786815,-0.00059656834,0.0056479387,-0.0059979255,-0.01671032,-0.0082330685,-0.0017544744,0.015202086,-0.008520787,-0.012766941,-0.021327354,-0.015333309,-0.092359275,-0.01666448,-0.012225122,-0.017390717,-0.001796707,-0.019986382,0.008724937,-0.038599875,-0.020174127,-0.0031538361,0.021196282,-0.042483598,-0.024682706,-0.040925145,0.024631003,0.009504316,0.0031550766,0.009888855,0.024441503,-0.023728615,-0.006822411,0.018813154,0.03925955,-0.036362596,-0.01741124,0.005582707,-0.014357929,0.016291248,-0.020017192,0.009013023,-0.0045064893,-0.12901062,0.0024711816,-0.004277638,-0.021809543,0.01818742,-0.0030945996,-0.0039685285,-0.00047468528,-0.011454802,0.012480421,0.0045565804,-0.026806146,-0.03371884,-0.008351203,-0.017261282,0.12352507,0.004733805,0.018316591,-0.027955448,-0.016978208,0.015733948,-0.020057276,0.0017250943,0.013125487,0.010372207,-0.030513816,0.03417878,-0.019050872,0.024113325,0.008826785,0.012053345,0.01908168,-0.011575329,-0.029001612,0.0123280175,-0.0055935,-0.008394542,-0.027964484,-0.0023019032,0.014253853,0.010574883,0.01130134,-0.0046410477,-0.032163654,-0.021113388,-0.009671858,-0.024442647,-0.02353686,-0.012336035,-0.0030777191,-0.019190105,-0.057718027,-0.0040828893,-0.014846871,0.03632439,0.0039261766,-0.016710745,0.007392858,0.028597567,0.011722327,0.021756439,-0.0061345864,-0.008863089,0.015819915,-0.03946997,-0.009653028,0.034666847,0.027184622,0.012236897,0.034685034,0.008641182,0.012861398,0.0041786064,-0.0055742143,0.01542257,-0.018065475,0.013981611,0.022281755,0.021352608,-0.019504642,-0.00026332279,-0.010980276,-0.011607256,-0.0017302535,0.013024999,0.014720719,-0.0045660757,0.0008873555,0.003029075,0.0018787639,0.0026973374,0.020745246,-0.006324673,0.001428839,-0.005348388,0.027119445,-0.0045236694,0.011729512,0.0011919197,-0.013083558,0.01922981,-0.015535042,0.02289776,0.012347919,0.007705927,0.04015183,-0.034339417,-0.01650744,0.03788245,-0.0039023736]",{"tags":39,"relatedLang":50,"relatedPosts":54},[40,42,44,46,48],{"name":14,"slug":41},"coding-agents",{"name":15,"slug":43},"open-models",{"name":13,"slug":45},"llm",{"name":17,"slug":47},"claude",{"name":16,"slug":49},"local-workflows",{"id":30,"slug":51,"title":52,"language":53},"5-shifts-in-llms-from-the-last-six-months-en","5 shifts in LLMs from the last six months","en",[55,61,67,73,79,85],{"id":56,"slug":57,"title":58,"cover_image":59,"image_url":59,"created_at":60,"category":29},"6f43934b-c0b5-4a6f-885d-526154b1ecdd","8-ai-coding-assistants-for-enterprise-teams-zh","8 個企業 AI 程式助手","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1779174267071-g3zi.png","2026-05-19T07:03:36.1425+00:00",{"id":62,"slug":63,"title":64,"cover_image":65,"image_url":65,"created_at":66,"category":29},"8395d8bd-7bfd-4aaf-9c18-59033a10ea87","why-wembanyama-game-3-should-change-spurs-expectations-zh","為什麼文班亞馬的第三戰表現，該改變馬刺對他的期待","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1779171232365-i7iz.png","2026-05-19T06:13:23.880797+00:00",{"id":68,"slug":69,"title":70,"cover_image":71,"image_url":71,"created_at":72,"category":29},"68a6ddb5-75b5-4581-9737-eafe2f718bd4","spurs-vs-timberwolves-game-5-takeaways-zh-tw-zh","馬刺 vs 灰狼 第五戰重點","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1779170039610-qjqz.png","2026-05-19T05:53:35.552606+00:00",{"id":74,"slug":75,"title":76,"cover_image":77,"image_url":77,"created_at":78,"category":29},"6ef82d1a-cd63-4531-9399-5367995ffb3c","wembanyama-game-5-revenge-test-wolves-zh","文班亞馬 Game 5 反擊戰","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1779169436870-egzm.png","2026-05-19T05:43:31.672533+00:00",{"id":80,"slug":81,"title":82,"cover_image":83,"image_url":83,"created_at":84,"category":29},"86ce3639-09f3-4b6f-aea0-30d588daab8d","5-reasons-webassembly-fits-edge-data-processing-zh","5 個 WebAssembly 邊緣優勢","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1779168830285-84xq.png","2026-05-19T05:33:21.815565+00:00",{"id":86,"slug":87,"title":88,"cover_image":89,"image_url":89,"created_at":90,"category":29},"9490f35a-38a9-4006-a1cc-00f8da10f80a","fever-monique-billings-early-2026-impact-zh","Billings 2026 首秀，Fever 補到深度","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1779164034029-nxli.png","2026-05-19T04:13:26.978775+00:00",[92,97,102,107,112,117,122,127,132,137],{"id":93,"slug":94,"title":95,"created_at":96},"ee073da7-28b3-4752-a319-5a501459fb87","ai-in-2026-what-actually-matters-now-zh","2026 AI 真正重要的事","2026-03-26T07:09:12.008134+00:00",{"id":98,"slug":99,"title":100,"created_at":101},"83bd1795-8548-44c9-9a7e-de50a0923f71","trump-ai-framework-power-speech-state-preemption-zh","川普 AI 框架瞄準電力、言論與州權","2026-03-26T07:12:18.695466+00:00",{"id":103,"slug":104,"title":105,"created_at":106},"ea6be18b-c903-4e54-97b7-5f7447a612e0","nvidia-gtc-2026-big-ai-announcements-zh","NVIDIA GTC 2026 重點拆解","2026-03-26T07:14:26.62638+00:00",{"id":108,"slug":109,"title":110,"created_at":111},"4bcec76f-4c36-4daa-909f-54cd702f7c93","claude-users-spreading-out-and-getting-better-zh","Claude 用戶更分散，也更會用","2026-03-26T07:22:52.325888+00:00",{"id":113,"slug":114,"title":115,"created_at":116},"bd903b15-2473-4178-9789-b7557816e535","openclaw-raises-hard-question-for-ai-models-zh","OpenClaw 逼問 AI 模型價值","2026-03-26T07:24:54.707486+00:00",{"id":118,"slug":119,"title":120,"created_at":121},"eeac6b9e-ad9d-4831-8eec-8bba3f9bca6a","gap-google-gemini-checkout-fashion-search-zh","Gap 把結帳搬進 Gemini","2026-03-26T07:28:23.937768+00:00",{"id":123,"slug":124,"title":125,"created_at":126},"0740e53f-605d-4d57-8601-c10beb126f3c","google-pushes-gemini-transition-to-march-2026-zh","Google 把 Gemini 轉換延到 2026 年 3…","2026-03-26T07:30:12.825269+00:00",{"id":128,"slug":129,"title":130,"created_at":131},"e660d801-2421-4529-8fa9-86b82b066990","metas-llama-4-benchmark-scandal-gets-worse-zh","Meta Llama 4 分數風波又擴大","2026-03-26T07:34:21.156421+00:00",{"id":133,"slug":134,"title":135,"created_at":136},"183f9e7c-e143-40bb-a6d5-67ba84a3a8bc","accenture-mistral-ai-sovereign-enterprise-deal-zh","Accenture 攜手 Mistral AI 賣主權 AI","2026-03-26T07:38:14.818906+00:00",{"id":138,"slug":139,"title":140,"created_at":141},"191d9b1b-768a-478c-978c-dd7431a38149","mistral-ai-faces-its-hardest-year-yet-zh","Mistral AI 迎來最硬的一年","2026-03-26T07:40:23.716374+00:00"]