{"id":87721,"topic":"ai","source":"手机新浪网","title":"新浪网· - 手机新浪网","url":"https://www.sina.cn/news/detail/5345900533908922.html","url_hash":"5d38584f3cbe13ed3a393da69e7706a233a5e5b5","author":"","summary":"<a href=\"https://news.google.com/rss/articles/CBMiY0FVX3lxTE03OEE3T1B2Ri1GUklBS3dfTnp1c3FONENsR216OXNSWXQxWnRhdVRhck5PaGZKamgtdmh2SzA5Vk11N0x2engwSVd5LVJiVVdCRTVaWGwwcFpVUHFHZnFZR05Xdw?oc=5\" target=\"_blank\">新浪网·</a>&nbsp;&nbsp;<font color=\"#6f6f6f\">手机新浪网</font>","content":"#小米开源MiMoV2.6大模型#\n小米发布并开源了Xiaomi MiMo-V2.6系列大模型，这是目前最强的开源模型，MiMo-V2.6-Pro在 Artificial Analysis 综合智能指数中取得 46 分，超过 Kimi K3 和 Qwen3.8 Max。\n而且智能提升，价格不变。V2.6系列沿用 V2.5系列的API定价，超低成本将帕累托线硬生生往上抬了一大截，太强了[老师好]。\n这可不是刷榜，MiMo训练都是直播的，小米直播Mimo 2.6 AI 的强化学习，不到2天时间用掉了130万美元，每小时烧3万美元。Token消耗、训练指标、成本全公开，虽然烧钱，但效果也很好。大家可以多用用，拿真实任务测试一下。        \n发布于 上海","image_url":"http://wx2.sinaimg.cn/thumbnail/b94cf269gy1ihc9avfszjj20x317c1kx.jpg","lang":"zh","published_at":"2026-09-22T02:51:00+00:00","fetched_at":"2026-09-22T04:15:08+00:00","status":"read","starred":0,"extract_state":"ok","summary_auto":"#小米开源MiMoV2.6大模型#\n小米发布并开源了Xiaomi MiMo-V2.6系列大模型，这是目前最强的开源模型，MiMo-V2.6-Pro在 Artificial Analysis 综合智能指数中取得 46 分，超过 Kimi K3 和 Qwen3.8 Max。\n而且智能提升，价格不变。V2.6系列沿用 V2.5系列的API定价，超低成本将帕累托线硬生生往上抬了一大截，太强了[老师好]。\n这可不是刷榜，MiMo训练都是直播的，小米直播Mimo 2.6 AI 的强化学习，不到2天时间用掉了130万美元，每小时烧3万美元。Token消耗、训练指标、成本全公开，虽然烧钱，但效果也很好。大家可以多用用，拿真实任务测试一下。        \n发布于 上海","cluster_id":3549909,"extract_retries":0,"extract_error":null,"contract_version":"news_item.v1","format_contract_version":"news_item_formats.v1","dedup_url":"https://www.sina.cn/news/detail/5345900533908922.html","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 330 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":330,"summary_length":330,"usable_text_length":330,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":330,"summary_length":330}},"news_item":{"id":87721,"canonical_url":"https://www.sina.cn/news/detail/5345900533908922.html","source_url":"https://www.sina.cn/news/detail/5345900533908922.html","title":"新浪网· - 手机新浪网","source_name":"手机新浪网","author":null,"published_at":"2026-09-22T02:51:00+00:00","locale":"zh","topic":"ai","tags":[],"rss_summary":"<a href=\"https://news.google.com/rss/articles/CBMiY0FVX3lxTE03OEE3T1B2Ri1GUklBS3dfTnp1c3FONENsR216OXNSWXQxWnRhdVRhck5PaGZKamgtdmh2SzA5Vk11N0x2engwSVd5LVJiVVdCRTVaWGwwcFpVUHFHZnFZR05Xdw?oc=5\" target=\"_blank\">新浪网·</a>&nbsp;&nbsp;<font color=\"#6f6f6f\">手机新浪网</font>","full_text":"#小米开源MiMoV2.6大模型#\n小米发布并开源了Xiaomi MiMo-V2.6系列大模型，这是目前最强的开源模型，MiMo-V2.6-Pro在 Artificial Analysis 综合智能指数中取得 46 分，超过 Kimi K3 和 Qwen3.8 Max。\n而且智能提升，价格不变。V2.6系列沿用 V2.5系列的API定价，超低成本将帕累托线硬生生往上抬了一大截，太强了[老师好]。\n这可不是刷榜，MiMo训练都是直播的，小米直播Mimo 2.6 AI 的强化学习，不到2天时间用掉了130万美元，每小时烧3万美元。Token消耗、训练指标、成本全公开，虽然烧钱，但效果也很好。大家可以多用用，拿真实任务测试一下。        \n发布于 上海","excerpt":"#小米开源MiMoV2.6大模型#\n小米发布并开源了Xiaomi MiMo-V2.6系列大模型，这是目前最强的开源模型，MiMo-V2.6-Pro在 Artificial Analysis 综合智能指数中取得 46 分，超过 Kimi K3 和 Qwen3.8 Max。\n而且智能提升，价格不变。V2.6系列沿用 V2.5系列的API定价，超低成本将帕累托线硬生生往上抬了一大截，太强了[老师好]。\n这可不是刷榜，MiMo训练都是直播的，小米直播Mimo 2.6 AI 的强化学习，不到2天时间用掉了130万美元，每小时烧3万美元。Token消耗、训练指标、成本全公开，虽然烧钱，但效果也很好。大家可以多用用，拿真实任务测试一下。        \n发布于 上海","extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 330 characters.","diagnostics_url":"/api/diagnose?url=https%3A//www.sina.cn/news/detail/5345900533908922.html","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 330 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":330,"summary_length":330,"usable_text_length":330,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":330,"summary_length":330}}},"display_formats":["compact","card","full","digest_section","json"]},"daily_stack_record":{"title":"新浪网· - 手机新浪网","url":"https://www.sina.cn/news/detail/5345900533908922.html","summary":"#小米开源MiMoV2.6大模型#\n小米发布并开源了Xiaomi MiMo-V2.6系列大模型，这是目前最强的开源模型，MiMo-V2.6-Pro在 Artificial Analysis 综合智能指数中取得 46 分，超过 Kimi K3 和 Qwen3.8 Max。\n而且智能提升，价格不变。V2.6系列沿用 V2.5系列的API定价，超低成本将帕累托线硬生生往上抬了一大截，太强了[老师好]。\n这可不是刷榜，MiMo训练都是直播的，小米直播Mimo 2.6 AI 的强化学习，不到2天时间用掉了130万美元，每小时烧3万美元。Token消耗、训练指标、成本全公开，虽然烧钱，但效果也很好。大家可以多用用，拿真实任务测试一下。        \n发布于 上海","source":"手机新浪网","date":"2026-09-22T02:51:00+00:00","content":"#小米开源MiMoV2.6大模型#\n小米发布并开源了Xiaomi MiMo-V2.6系列大模型，这是目前最强的开源模型，MiMo-V2.6-Pro在 Artificial Analysis 综合智能指数中取得 46 分，超过 Kimi K3 和 Qwen3.8 Max。\n而且智能提升，价格不变。V2.6系列沿用 V2.5系列的API定价，超低成本将帕累托线硬生生往上抬了一大截，太强了[老师好]。\n这可不是刷榜，MiMo训练都是直播的，小米直播Mimo 2.6 AI 的强化学习，不到2天时间用掉了130万美元，每小时烧3万美元。Token消耗、训练指标、成本全公开，虽然烧钱，但效果也很好。大家可以多用用，拿真实任务测试一下。        \n发布于 上海","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//www.sina.cn/news/detail/5345900533908922.html","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 330 characters.","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 330 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":330,"summary_length":330,"usable_text_length":330,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":330,"summary_length":330}},"tags":[]},"fallback_formats":["markdown","json","html"],"actions":{"read":"/item/87721","export_markdown":"/api/items/87721/export?format=markdown","export_json":"/api/items/87721/export?format=json","diagnose":"/api/diagnose?url=https%3A//www.sina.cn/news/detail/5345900533908922.html"},"formats":{"full":{"id":87721,"title":"新浪网· - 手机新浪网","url":"https://www.sina.cn/news/detail/5345900533908922.html","source":"手机新浪网","author":null,"published_at":"2026-09-22T02:51:00+00:00","locale":"zh","topic":"ai","tags":[],"excerpt":"#小米开源MiMoV2.6大模型#\n小米发布并开源了Xiaomi MiMo-V2.6系列大模型，这是目前最强的开源模型，MiMo-V2.6-Pro在 Artificial Analysis 综合智能指数中取得 46 分，超过 Kimi K3 和 Qwen3.8 Max。\n而且智能提升，价格不变。V2.6系列沿用 V2.5系列的API定价，超低成本将帕累托线硬生生往上抬了一大截，太强了[老师好]。\n这可不是刷榜，MiMo训练都是直播的，小米直播Mimo 2.6 AI 的强化学习，不到2天时间用掉了130万美元，每小时烧3万美元。Token消耗、训练指标、成本全公开，虽然烧钱，但效果也很好。大家可以多用用，拿真实任务测试一下。        \n发布于 上海","full_text":"#小米开源MiMoV2.6大模型#\n小米发布并开源了Xiaomi MiMo-V2.6系列大模型，这是目前最强的开源模型，MiMo-V2.6-Pro在 Artificial Analysis 综合智能指数中取得 46 分，超过 Kimi K3 和 Qwen3.8 Max。\n而且智能提升，价格不变。V2.6系列沿用 V2.5系列的API定价，超低成本将帕累托线硬生生往上抬了一大截，太强了[老师好]。\n这可不是刷榜，MiMo训练都是直播的，小米直播Mimo 2.6 AI 的强化学习，不到2天时间用掉了130万美元，每小时烧3万美元。Token消耗、训练指标、成本全公开，虽然烧钱，但效果也很好。大家可以多用用，拿真实任务测试一下。        \n发布于 上海","reading_time_min":1,"extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 330 characters.","diagnostics_url":"/api/diagnose?url=https%3A//www.sina.cn/news/detail/5345900533908922.html","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 330 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":330,"summary_length":330,"usable_text_length":330,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":330,"summary_length":330}}},"quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 330 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":330,"summary_length":330,"usable_text_length":330,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":330,"summary_length":330}},"actions":{"read":"/item/87721","export_markdown":"/api/items/87721/export?format=markdown","export_json":"/api/items/87721/export?format=json","diagnose":"/api/diagnose?url=https%3A//www.sina.cn/news/detail/5345900533908922.html"}},"digest":{"id":87721,"title":"新浪网· - 手机新浪网","url":"https://www.sina.cn/news/detail/5345900533908922.html","source":"手机新浪网","topic":"ai","published_at":"2026-09-22T02:51:00+00:00","excerpt":"#小米开源MiMoV2.6大模型# 小米发布并开源了Xiaomi MiMo-V2.6系列大模型，这是目前最强的开源模型，MiMo-V2.6-Pro在 Artificial Analysis 综合智能指数中取得 46 分，超过 Kimi K3 和 Qwen3.8 Max。 而且智能提升，价格不变。V2.6系列沿用 V2.5系列的API定价，超低成本将帕累托线硬生生往上抬了一大截，太强了[老师好]。 这可不是刷榜，MiMo训练都是直播的，小米直播Mimo 2.6 AI…","quality_bucket":"high","quality_reason":"High confidence: full text extraction produced 330 characters.","reading_time_min":1,"cluster_id":3549909},"card":{"display_title":"新浪网· - 手机新浪网","subtitle":"手机新浪网 · 2026-09-22","summary":"#小米开源MiMoV2.6大模型# 小米发布并开源了Xiaomi MiMo-V2.6系列大模型，这是目前最强的开源模型，MiMo-V2.6-Pro在 Artificial Analysis 综合智能指数中取得 46 分，超过 Kimi K3 和 Qwen3.8 Max。 而且智能提升，价格不变。V2.6系列沿用 V2.5系列的API定价，超低成本将帕累托线硬生生往上抬了一大截，太强了[老师好]。…","badges":["quality:high"],"links":{"read":"/item/87721","original":"https://www.sina.cn/news/detail/5345900533908922.html","diagnose":"/api/diagnose?url=https%3A//www.sina.cn/news/detail/5345900533908922.html"},"quality_warning":null},"export":{"title":"新浪网· - 手机新浪网","url":"https://www.sina.cn/news/detail/5345900533908922.html","summary":"#小米开源MiMoV2.6大模型#\n小米发布并开源了Xiaomi MiMo-V2.6系列大模型，这是目前最强的开源模型，MiMo-V2.6-Pro在 Artificial Analysis 综合智能指数中取得 46 分，超过 Kimi K3 和 Qwen3.8 Max。\n而且智能提升，价格不变。V2.6系列沿用 V2.5系列的API定价，超低成本将帕累托线硬生生往上抬了一大截，太强了[老师好]。\n这可不是刷榜，MiMo训练都是直播的，小米直播Mimo 2.6 AI 的强化学习，不到2天时间用掉了130万美元，每小时烧3万美元。Token消耗、训练指标、成本全公开，虽然烧钱，但效果也很好。大家可以多用用，拿真实任务测试一下。        \n发布于 上海","source":"手机新浪网","date":"2026-09-22T02:51:00+00:00","content":"#小米开源MiMoV2.6大模型#\n小米发布并开源了Xiaomi MiMo-V2.6系列大模型，这是目前最强的开源模型，MiMo-V2.6-Pro在 Artificial Analysis 综合智能指数中取得 46 分，超过 Kimi K3 和 Qwen3.8 Max。\n而且智能提升，价格不变。V2.6系列沿用 V2.5系列的API定价，超低成本将帕累托线硬生生往上抬了一大截，太强了[老师好]。\n这可不是刷榜，MiMo训练都是直播的，小米直播Mimo 2.6 AI 的强化学习，不到2天时间用掉了130万美元，每小时烧3万美元。Token消耗、训练指标、成本全公开，虽然烧钱，但效果也很好。大家可以多用用，拿真实任务测试一下。        \n发布于 上海","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//www.sina.cn/news/detail/5345900533908922.html","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 330 characters.","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 330 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":330,"summary_length":330,"usable_text_length":330,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":330,"summary_length":330}},"tags":[],"format_contract_version":"news_item_formats.v1"}}}