[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-xiaomi-mimo-v2-5-five-inference-optimization":3,"topics-all":36,"news-related-b4a4434f-6b1e-4ff3-ae69-2d9e82ab3e29":55},{"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},"b4a4434f-6b1e-4ff3-ae69-2d9e82ab3e29","小米MiMo-V2.5首度揭秘：五大推理优化技术如何实现「降价不亏本」","小米MiMo大模型团队近日首次系统披露了MiMo-V2.5系列API永久降价背后的技术路径。外界看到的是一次次「骨折价」，团队要解决的却是如何在降价后依然维持收支平衡这道难题。\n\nV2.5版本实现了五大核心突破。**KVCache双池+SWA-aware前缀树**解决了长prompt场景下的缓存碎片化问题，将前缀复用率显著提升；**GCache分布式缓存**则在跨请求层面做共享，减少重复计算。**KVCache亲和调度**根据请求特征动态分配缓存资源，提升显存利用率。\n\n在Decode阶段，团队引入了**MTP（Multi-Token Prediction）加速**，一次推理可输出多个token，直接提升吞吐量。**多模态推理优化**则针对图像编码路径做了专门加速，降低端到端延迟。\n\n从实现路径看，小米走的是一条「软硬协同优化」路线——不依赖单点突破，而是从缓存策略、调度逻辑到模型结构全链路协同。这也解释了为何V2.5能在保持效果的同时支撑起更低的价格。\n\n值得关注的是，这套优化方案并不依赖特殊硬件，正是因为如此，MiMo才能在降价后依然保持商业可持续。对行业而言，这种「工程密集型降本」路径比单纯靠硬件红利或压缩参数更能持续。\n\n小米同时启动了「百万亿Token创造者激励计划」，目前已有超54万开发者申请，累计发放100万亿免费Token。这一规模说明降价策略已真正触达用户，而非单纯的市场噱头。团队下一步的方向，应该是让这些技术优化在生产环境中的持续验证。","https:\u002F\u002F36kr.com\u002Fnewsflashes\u002F3832525007284097","5e4fd3d1-9cb4-44a6-bae5-9ffb449c05c1",[10,14,17,20],{"id":11,"name":12,"slug":12,"description":13,"color":13},"e676a5cf-1f24-472f-a765-86fa21a1bc3c","ai-model",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},"01598627-1ea6-4b27-a5d8-874971571a71","llm",{"id":21,"name":22,"slug":22,"description":13,"color":13},"b9bd9039-fcdb-41a8-b85b-fc1587def2b9","open-source",[24],{"id":25,"lang":26,"title":27,"summary":28,"content":13},"51872cad-559f-40b9-839f-680961fe154d","en","Inside Xiaomi MiMo-V2.5: five tricks for cheaper inference","36Kr published a deep dive into Xiaomi's MiMo-V2.5 inference optimization techniques. The five techniques — speculative decoding, KV cache compression, MoE expert scheduling, dynamic batching, and quantization — combine to reduce inference cost by 60% while maintaining quality, allowing Xiaomi to offer aggressive pricing without margin loss.","xiaomi-mimo-v2-5-five-inference-optimization","2026-05-31T04:00:00Z","2026-05-31T04:06:41.900171Z","2026-08-19T02:08:40.142862Z",true,"agent",140,[37,46],{"slug":38,"tag_slug":38,"title_zh":39,"title_en":40,"intro_zh":41,"intro_en":42,"id":43,"is_active":33,"created_at":44,"modified_at":45},"ai-for-science","AI for Science 2026：从 UniPert 到 GPT-Rosalind 的硬核进化","AI for Science 2026: from UniPert to GPT-Rosalind","生命科学、化学材料、物理世界模型——AI 正在从\"语言工具\"变成\"实验伙伴\"。本专题收录 AI 在三大科学方向的关键节点：UniPert 统一基因与化学扰动空间、GPT-Rosalind 端到端生命科学推理、达摩院 AI 智能体 28 小时找到 4 种超导新材料、Anthropic Claude Science 把工作台做成标准品。","From language tool to lab partner — AI is reshaping life sciences, chemistry\u002Fmaterials, and physical world models. This topic covers the key milestones: UniPert unifying genetic-chemical perturbation spaces, GPT-Rosalind's end-to-end life-sciences reasoning, DAMO's AI agent discovering 4 superconducting materials in 28 hours, and Anthropic's Claude Science workbench going mainstream.","988a4300-5fab-41c4-b5d8-63711a2dc757","2026-09-10T01:34:15.296649Z","2026-09-10T01:34:15.296663Z",{"slug":47,"tag_slug":47,"title_zh":48,"title_en":49,"intro_zh":50,"intro_en":51,"id":52,"is_active":33,"created_at":53,"modified_at":54},"h3-series","MiniMax H3 系列：从开源权重到 35 倍吞吐","MiniMax H3 Series: from open weights to 35x throughput","MiniMax H3 自 2026 年 8 月开源以来节奏密集：官方把生成、参考与编辑收回一个模型；ComfyUI 当天压进 RTX 3060；摩尔线程 3 小时完成国产 GPU 适配；fal 后训练版把吞吐拉到 35 倍；FastH3 蒸馏再砍推理成本。本专题持续追踪 H3 的发布—开源—蒸馏—部署全链路。","Since MiniMax open-sourced H3 in August 2026 the pace has been relentless: one unified omni-modal model, same-day ComfyUI support down to an RTX 3060, a 3-hour Day-0 port to Moore Threads GPUs, fal's post-trained H3 Max at 35x throughput, and FastH3 distillation cutting inference cost further. This topic tracks the full H3 chain — release, open weights, distillation, deployment.","83ef0daa-3c31-4cb3-86ed-e5ee58654d5f","2026-09-08T07:33:19.942193Z","2026-09-08T07:33:19.942209Z",{"items":56},[57,62,67,72,77,82],{"id":58,"title":59,"news_slug":60,"published_at":61},"e6865eab-e2f9-451e-9823-8c336e93452a","小米MiMo-V2.5-Pro开源：万亿参数MoE+1M上下文，长程Agent能力新突破","xiaomi-mimo-v2-5-pro-1t-moe-672-toolcalls","2026-05-10T11:10:00+00:00",{"id":63,"title":64,"news_slug":65,"published_at":66},"28c41f06-d20f-481c-b133-cd109af3aed1","答对之后停不下来:微软团队揪出在线蒸馏的 EOS 错配元凶","eos-mismatch-opd-length-inflation","2026-09-18T21:09:06+00:00",{"id":68,"title":69,"news_slug":70,"published_at":71},"4497a0c5-e9b8-42d5-8d72-cfa0a49c1fba","Mistral Small 4 加入 Firefox Smart Window：开放权重模型第一次进浏览器助手默认菜单","mistral-mozilla-firefox-smart-window-moe","2026-09-17T19:00:00+00:00",{"id":73,"title":74,"news_slug":75,"published_at":76},"2e27016d-b90e-45c7-825a-41fd1e435c80","JHU 新研究:组合持续学习机制,百任务记忆留存从 1.2% 提到 34.9%","compose-cl-long-horizon-memorization","2026-09-16T15:10:00+00:00",{"id":78,"title":79,"news_slug":80,"published_at":81},"30fca629-bace-4832-9789-b44aa8c8989d","学生团队从零训出开源 7B 模型 ZGCM-1:数学推理硬刚 235B 前沿","zgcm-1-open-7b-foundation-model","2026-09-15T19:10:00+00:00",{"id":83,"title":84,"news_slug":85,"published_at":86},"2731ed1c-17c3-4d85-9174-983cf50743e3","地铁售票机上的 AI 大考:2.6GB 端侧模型 91.32 分超 GPT-5.6,规则基线也拿 84.6","metrollm-bench-transit-kiosk-llm","2026-09-12T23:08:18+00:00"]