[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-grok-4-3-xai-207-tps-cheap-fast":3,"news-related-e2a935d5-4893-4acb-bdb5-1783c19eeb20":36},{"id":4,"title":5,"summary":6,"content":6,"original_url":7,"source_id":8,"tags":9,"translations":23,"news_slug":29,"published_at":30,"created_at":31,"modified_at":32,"is_published":33,"publish_type":34,"image_url":13,"view_count":35},"e2a935d5-4893-4acb-bdb5-1783c19eeb20","xAI悄然发布Grok 4.3：速度致胜，但智能仍未登顶","xAI悄然发布Grok 4.3：速度致胜，但智能仍未登顶\n\n4月30日，xAI悄悄在API上线了Grok 4.3——没有发布会，没有 Elon 发推预告，只有开发控制台里一条小小的迁移提示。这一反常态的低调发布，本身就是一个信号。\n\n**技术参数与定价**\n\nGrok 4.3的官方定价为输入每百万token 1.25美元、输出每百万token 2.5美元，上下文窗口达100万token，输出速度可达207 token\u002F秒。这些数字看起来相当激进——比GPT-5.5便宜一半以上，比Claude Opus 4.7便宜约75%。\n\n**真实跑分：不功不过**\n\n然而，独立基准测试给出的Intelligence Index得分为53，位列当前三大前沿模型之末：GPT-5.5得60，Claude Opus 4.7得57，Gemini 3.1 Pro Preview同获57。这个7分的差距，在多步推理、代码生成和验证类任务上，意味着明显的体验落差。\n\n这实际上反映了xAI战略重心的一次重要转移。Grok 4.3不再试图成为最聪明的模型，而是在速度与成本这个维度建立竞争优势。对于需要实时响应、低延迟的语音代理、直播对话、长篇内容生成等场景，207 token\u002F秒的输出速度是切实的生产力工具。\n\n**容易被忽视的代价**\n\n但Grok 4.3有一个隐藏弱点：Time-to-First-Token（TTFT，首token等待时间）高达12.65秒，而同价位模型的中间值仅为2.82秒。这意味着Grok 4.3需要更长的思考时间才能开始输出，对于短查询反而比GPT-5.5更慢——速度优势只在生成大量内容时才真正体现。\n\nxAI官方文档倒是很诚实：他们没有将Grok 4.3定位为最聪明，而是最快、最便宜的高端级模型。这个低调的定位，或许才是它真正的产品定义。\n\n**所以呢**\n\nGrok 4.3是一枚精准的生产力工具，而非通用智能的挑战者。xAI在追逐OpenAI\u002FAnthropic的智能巅峰和成为最快最便宜的效率选择之间，选择了后者。这个战略是否正确，取决于市场到底需要什么样的xAI——而这一点，或许连Elon自己也在摸索。","https:\u002F\u002Fwww.roborhythms.com\u002Fgrok-4-3-release-april-2026\u002F","b82e17a3-1dbd-4b5d-88dc-9f518f917cc0",[10,14,17,20],{"id":11,"name":12,"slug":12,"description":13,"color":13},"120fa59a-ff6f-4537-9bf5-f818df636a0e","benchmark",null,{"id":15,"name":16,"slug":16,"description":13,"color":13},"0ef8513a-0a26-42f0-b6f9-5b6dadded45c","efficiency",{"id":18,"name":19,"slug":19,"description":13,"color":13},"0a93ec8e-ea39-4693-81de-563ca8c173f7","inference",{"id":21,"name":22,"slug":22,"description":13,"color":13},"01598627-1ea6-4b27-a5d8-874971571a71","llm",[24],{"id":25,"lang":26,"title":27,"summary":28,"content":13},"86c6f80e-c063-413f-b523-50d1e0d6a672","en","Grok 4.3 quietly arrives: speed wins, smarts still lag","xAI quietly released Grok 4.3 in its API on April 30 — no launch event, no Elon tweet teaser, just a small migration note in the developer console. This unusually low-key release is itself a signal.\n\n**Technical parameters and pricing**\n\nGrok 4.3's official pricing is $1.25 per million input tokens, $2.50 per million output tokens, with a 1-million-token context window, and output speed up to 207 tokens\u002Fsecond. These numbers look quite aggressive — over 50% cheaper than GPT-5.5, about 75% cheaper than Claude Opus 4.7.\n\n**Real benchmark: neither here nor there**\n\nBut independent benchmark testing gives an Intelligence Index score of 53, ranking last among the three top frontier models: GPT-5.5 scores 60, Claude Opus 4.7 scores 57, Gemini 3.1 Pro Preview also gets 57. This 7-point gap, in multi-step reasoning, code generation, and verification tasks, means a clear experience divide.\n\nThis actually reflects an important shift in xAI's strategic focus. Grok 4.3 is no longer trying to be the smartest model, but is building a competitive advantage in the speed-and-cost dimension. For scenarios requiring real-time response, low-latency voice agents, livestream dialogue, long-form content generation, etc., 207 tokens\u002Fsecond output speed is a real productivity tool.\n\n**An easily-overlooked cost**\n\nBut Grok 4.3 has a hidden weakness: Time-to-First-Token (TTFT) as high as 12.65 seconds, while the median for same-priced models is only 2.82 seconds. This means Grok 4.3 needs much longer thinking time before it starts outputting, and for short queries it's actually slower than GPT-5.5 — the speed advantage only truly manifests when generating large volumes of content.\n\nxAI's official docs are honest: they don't position Grok 4.3 as the smartest, but as the fastest, cheapest high-end model. This low-key positioning may actually be its real product definition.\n\n**So what**\n\nGrok 4.3 is a precision productivity tool, not a challenger for general intelligence. xAI, choosing between chasing OpenAI\u002FAnthropic's intelligence peak and becoming the fastest and cheapest efficiency option, chose the latter. Whether this strategy is right depends on what kind of xAI the market actually needs — and that, perhaps even Elon himself is still figuring out.","grok-4-3-xai-207-tps-cheap-fast","2026-05-03T16:01:00Z","2026-05-03T16:05:56.033132Z","2026-08-19T02:08:40.142862Z",true,"agent",113,{"items":37},[38,43,48,53,58,63],{"id":39,"title":40,"news_slug":41,"published_at":42},"7cacddc6-fa02-4de9-84a2-c3320e225571","因果归因剪枝 CAP：让 LLM 推理能力不再随稀疏化而流失","cap-causal-attribution-pruning-arc-61pct","2026-06-20T22:14:08.915874+00:00",{"id":44,"title":45,"news_slug":46,"published_at":47},"9a1e1c85-60eb-47c6-92b5-bace1746e217","大模型竞争进入下半场：从「比参数」到「比部署」——2026年5月技术格局观察","llm-2nd-half-deploy-vs-params-may-2026","2026-05-25T05:15:00+00:00",{"id":49,"title":50,"news_slug":51,"published_at":52},"7dbe12ab-8a86-4e19-a849-b6b0be3f985c","Qwen3.7-Max评测揭示推理代价：97M token输出背后的效率博弈","qwen3-7-max-97m-tokens-extended-thinking","2026-05-22T10:00:00+00:00",{"id":54,"title":55,"news_slug":56,"published_at":57},"9bb023ae-147a-4081-a973-5638e260803f","1M 上下文实测：Gemini 3.1 Pro 与 Opus 4.7 稳，GPT-5.5 在 512K 衰减","1m-context-multihop-benchmark-cliff-degradation","2026-05-15T22:00:00+00:00",{"id":59,"title":60,"news_slug":61,"published_at":62},"72a30e44-f38d-42af-af4a-32d265f76608","EfficientLLM：大模型效率研究的首次系统性「全景扫描」","efficient-llm-benchmark-panorama-tradeoff","2026-05-14T08:10:00+00:00",{"id":64,"title":65,"news_slug":66,"published_at":67},"0620b8c4-65be-4230-8adb-956c282bdc8b","DeepSeek V4-Pro 代码能力跃升至第三：压缩注意力机制如何重写百万级上下文效率","deepseek-v4-pro-csa-hca-1456-elo-27pct-flops","2026-04-27T01:00:00+00:00"]