[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"tag-context-window":3},{"tag":4,"articles":11,"peer_article_count":73},{"id":5,"name":6,"slug":7,"article_count":8,"description_zh":9,"description_en":10},"c8736be8-03af-4947-bea5-59ee9ef43de3","context window","context-window",5,"context window 指的是模型一次能讀進並參考多少文字、程式碼或對話內容，直接影響長文件分析、程式除錯、代理記憶與多輪工作流程。它牽涉到 token 上限、成本、延遲與資訊遺失的取捨。","Context window is the amount of text, code, or conversation a model can hold at once. It shapes long-document analysis, debugging, agent memory, and multi-step workflows, while forcing trade-offs between token limits, cost, latency, and retained state.",[12,21,29,36,43,51,58,65],{"id":13,"slug":14,"title":15,"summary":16,"category":17,"image_url":18,"cover_image":18,"language":19,"created_at":20},"7fa092e5-5be7-41b6-bb60-2792c0a79fac","gpt-56-rumors-2m-context-coding-gains-en","GPT-5.6 rumors point to 2M context and coding gains","Rumors point to GPT-5.6 and GPT-5.6 Pro arriving June 25 with 2M context, stronger coding agents, and lower prices than rivals.","model-release","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782248567152-pbak.png","en","2026-06-23T21:02:23.553973+00:00",{"id":22,"slug":23,"title":24,"summary":25,"category":26,"image_url":27,"cover_image":27,"language":19,"created_at":28},"0802f58b-dd51-4bae-8881-4f873ed99eb0","gpt-56-fix-and-upgrade-release-en","GPT-5.6 looks like a fix-and-upgrade release","5 things GPT-5.6 may change, from a late-June launch window to a rumored 1.5M-token context and faster coding.","industry","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1781756270916-jh5e.png","2026-06-18T04:17:28.410175+00:00",{"id":30,"slug":31,"title":32,"summary":33,"category":17,"image_url":34,"cover_image":34,"language":19,"created_at":35},"768b8352-3611-4527-99b1-dd1c87c45fea","claude-opus-4-8-api-pricing-benchmarks-openrouter-en","Claude Opus 4.8: $5\u002F$25 API pricing, 1M context","OpenRouter lists Anthropic’s Claude Opus 4.8 at $5 per million input tokens and $25 per million output tokens with a 1M-token context window.","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1781086675114-cu4j.png","2026-06-10T10:17:25.12768+00:00",{"id":37,"slug":38,"title":39,"summary":40,"category":17,"image_url":41,"cover_image":41,"language":19,"created_at":42},"160cf218-8ea5-44d3-b250-5fc8f8b25b73","what-we-know-about-gpt-56-release-date-en","What We Know About GPT-5.6's Release Date","OpenAI has not announced GPT-5.6, but hiring, infrastructure work, and model rumors point to a late-2024 or early-2025 window.","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1780574580198-szkr.png","2026-06-04T12:02:35.698162+00:00",{"id":44,"slug":45,"title":46,"summary":47,"category":48,"image_url":49,"cover_image":49,"language":19,"created_at":50},"0617c502-198c-449e-a9e5-1c28b590e062","why-pi-mcp-adapter-is-the-right-way-to-use-mcp-en","Why Pi MCP Adapter Is the Right Way to Use MCP","Pi MCP Adapter is the right way to use MCP because it cuts token waste without giving up useful tools.","tools","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1779034424252-f5wh.png","2026-05-17T16:13:20.396868+00:00",{"id":52,"slug":53,"title":54,"summary":55,"category":17,"image_url":56,"cover_image":56,"language":19,"created_at":57},"cb45188a-2e6e-4ac7-95f0-39cbd2f7d7a2","gpt-5-4-benchmarks-2026-scores-rankings-en","GPT-5.4 Scores 97.6 in Knowledge Benchmarks","GPT-5.4 tops knowledge benchmarks with 97.6, ranks #2 overall on BenchLM, and posts a 1.05M-token context window.","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1776082204490-nq2r.png","2026-04-13T12:09:40.792366+00:00",{"id":59,"slug":60,"title":61,"summary":62,"category":48,"image_url":63,"cover_image":63,"language":19,"created_at":64},"1e2b9ee3-510a-4d07-855a-766d0be9874c","claude-code-setup-guide-researchers-en","Claude Code Setup Guide for Research Workflows","A practical setup guide for Claude Code in research workflows, with terminal tips, context-window advice, and pricing details.","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1775058381471-uefp.png","2026-04-01T13:15:40.740947+00:00",{"id":66,"slug":67,"title":68,"summary":69,"category":70,"image_url":71,"cover_image":71,"language":19,"created_at":72},"01299403-0ffd-4a04-abbb-5b4d792fd01c","agent-memory-framework-analysis-en","Agent Memory: How AI Agents Keep State","Agent memory lets AI agents retain state across tasks. Here’s how short-, long-, and external memory shape real agent systems.","ai-agent","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1775058027759-90qv.png","2026-04-01T10:21:33.504368+00:00",8]