{"id":51288,"date":"2026-03-19T12:17:02","date_gmt":"2026-03-19T04:17:02","guid":{"rendered":"https:\/\/www.1ai.net\/?p=51288"},"modified":"2026-03-19T12:17:02","modified_gmt":"2026-03-19T04:17:02","slug":"%e5%b0%8f%e7%b1%b3%e6%b7%b1%e5%a4%9c%e4%b8%8a%e7%ba%bf%e4%b8%89%e5%a4%a7%e8%87%aa%e7%a0%94-mimo-v2-%e7%b3%bb%e5%88%97%e6%a8%a1%e5%9e%8b","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/51288.html","title":{"rendered":"Three self-research MiMo-V2 series models for rice at night"},"content":{"rendered":"

3 \u6708 19 \u65e5\u6d88\u606f\uff0c\u4eca\u5929\u51cc\u6668\uff0c\u5c0f\u7c73<\/a>\u96c6\u4e2d\u53d1\u5e03\u4e86 MiMo V2 \u7cfb\u5217\u4e09\u6b3e\u6a21\u578b<\/a>\uff1a\u65d7\u8230\u8bed\u8a00\u57fa\u5ea7 MiMo-V2-Pro\u3001\u5168\u6a21\u6001 Agent \u57fa\u5ea7 MiMo-V2-Omni\uff0c\u4ee5\u53ca\u8bed\u97f3\u5408\u6210\u6a21\u578b MiMo-V2-TTS\u3002<\/p>\n

\"2d88ec6fj00tc4p6x00end000u000o9p\"<\/p>\n

MiMo-V2-Pro\uff1a<\/p>\n

\u603b\u53c2\u6570\u91cf\u8d85\u8fc7 1T\uff0c\u6fc0\u6d3b\u53c2\u6570\u4e3a 42B\uff0c\u91c7\u7528\u6df7\u5408\u6ce8\u610f\u529b\uff08Hybrid Attention\uff09\u67b6\u6784\uff0c\u6df7\u5408\u6bd4\u4f8b\u4ece\u524d\u4ee3\u7684 5:1 \u63d0\u5347\u81f3 7:1\uff0c\u5e76\u652f\u6301 1M \u8d85\u957f\u4e0a\u4e0b\u6587\u7a97\u53e3\uff1b<\/p>\n

\u5728 OpenClaw\u3001Claude Code \u7b49\u667a\u80fd\u4f53\u6846\u67b6\u4e2d\uff0c\u53ef\u5728\u65e0\u4eba\u5de5\u5e72\u9884\u6761\u4ef6\u4e0b\u5b8c\u6210\u590d\u6742\u5de5\u4f5c\u6d41\u7f16\u6392\u3001\u957f\u7a0b\u89c4\u5212\u4e0e\u7cbe\u51c6\u5de5\u5177\u8c03\u7528\uff1b<\/p>\n

Coding \u80fd\u529b\u65b9\u9762\uff0c\u5c0f\u7c73\u5185\u90e8\u5de5\u7a0b\u5e08\u8bc4\u6d4b\u4f53\u611f\u63a5\u8fd1 Claude Opus 4.6\uff0c\u5177\u5907\u66f4\u5f3a\u7684\u7cfb\u7edf\u8bbe\u8ba1\u4e0e\u4efb\u52a1\u89c4\u5212\u80fd\u529b\uff1b<\/p>\n

API \u5b9a\u4ef7\u4e3a Claude \u540c\u7ea7\u6a21\u578b\u7684\u7ea6 1\/5\uff0c256K \u4e0a\u4e0b\u6587\u4ee5\u5185\u8f93\u5165 $1 \/ \u767e\u4e07 token\uff0c\u8f93\u51fa $3 \/ \u767e\u4e07 token\uff1b1M \u4e0a\u4e0b\u6587\u4ee5\u5185\u8f93\u5165 $2 \/ \u767e\u4e07 token\uff0c\u8f93\u51fa $6 \/ \u767e\u4e07 token\u3002<\/p>\n

MiMo-V2-Omni \u5219\u662f\u5c0f\u7c73\u9996\u4e2a\u5728\u57fa\u5ea7\u5c42\u9762\u7edf\u4e00\u611f\u77e5\u4e0e\u884c\u52a8\u7684\u5168\u6a21\u6001\u6a21\u578b\uff0c\u878d\u5408\u6587\u672c\u3001\u89c6\u89c9\u4e0e\u8bed\u97f3\u8f93\u5165\uff0c\u652f\u6301 256K \u4e0a\u4e0b\u6587\uff0c\u5b9a\u4ef7\u4e3a\u8f93\u5165 $0.4 \/ \u767e\u4e07 token\uff0c\u8f93\u51fa $2 \/ \u767e\u4e07 token\u3002<\/p>\n

\u5b98\u65b9\u5ba3\u79f0\u5176\u97f3\u9891\u7406\u89e3\u7efc\u5408\u8868\u73b0\u8d85\u8d8a Gemini 3 Pro\uff0c\u56fe\u50cf\u7406\u89e3\u8d85\u8d8a Claude Opus 4.6\uff1b\u5728\u5b9e\u9645\u5e94\u7528\u573a\u666f\u4e2d\uff0cMiMo-V2-Omni \u53ef\u7ed3\u5408 OpenClaw \u6846\u67b6\u5b9e\u73b0\u6d4f\u89c8\u5668\u81ea\u52a8\u5316\u64cd\u4f5c\uff0c\u5305\u62ec\u8de8\u5e73\u53f0\u6bd4\u4ef7\u4e0b\u5355\u3001\u77ed\u89c6\u9891\u5236\u4f5c\u4e0e\u53d1\u5e03\u7b49\u7aef\u5230\u7aef\u4efb\u52a1\u3002<\/p>\n

MiMo-V2-TTS \u8bed\u97f3\u5408\u6210\u6a21\u578b\u652f\u6301\u4ece\u6574\u4f53\u98ce\u683c\u5230\u53e5\u5185\u5c40\u90e8\u60c5\u7eea\u7684\u591a\u7c92\u5ea6\u63a7\u5236\uff0c\u80fd\u591f\u81ea\u7136\u5904\u7406\u6807\u70b9\u3001\u8bed\u6c14\u8bcd\u7b49\u683c\u5f0f\u4fe1\u53f7\uff0c\u5e76\u652f\u6301\u4e1c\u5317\u8bdd\u3001\u56db\u5ddd\u8bdd\u3001\u7ca4\u8bed\u7b49\u591a\u79cd\u65b9\u8a00\uff0c\u4ee5\u53ca\u89d2\u8272\u626e\u6f14\u5f0f\u98ce\u683c\u5316\u6f14\u7ece\u548c\u6b4c\u58f0\u5408\u6210\u3002<\/p>\n

\u4e09\u6b3e\u6a21\u578b\u5747\u5df2\u63a5\u5165\u91d1\u5c71 WPS \u7075\u7280\uff0cMiMo-V2-Pro \u540c\u6b65\u4e0a\u7ebf\u5c0f\u7c73\u624b\u673a\u7aef Agent \u4ea7\u54c1 miclaw \u53ca\u5c0f\u7c73\u6d4f\u89c8\u5668\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"

On March 19, in the early morning hours of this morning, Mimo V2 focused on three models: the flagship language base MiMo-V2-Pro, the full-mod Agent base MiMo-V2-Omni, and the voice synthesis model MiMo-V2-TTS. MiMo-V2-Pro: Total parameter mass exceeding 1T, activated parameter 42B, using the Hybrid Attention structure, with a mixed ratio of 5:1 up to 7:1 in the previous generation, and support for a 1M super-long context window<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[146],"tags":[1114,1489],"collection":[],"class_list":["post-51288","post","type-post","status-publish","format-standard","hentry","category-news","tag-1114","tag-1489"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/51288","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=51288"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/51288\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=51288"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=51288"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=51288"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=51288"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}