{"id":"8f5228bd5d9a","page":"/c/8f5228bd5d9a/","topic":"LLM performance across six benchmarks","publisher":"Z.ai","official":true,"platform":"X","publishedAt":"2026-08-26T14:12:36.000Z","model":"GLM-5","form":"grouped-bar","metrics":{"views":6958294,"likes":23834,"reposts":2615,"quotes":2012,"replies":1002,"saves":4188},"readAt":"2026-10-13T17:43:55.213Z","chartsInPost":1,"score":78,"scoreWeights":{"views":0.4,"likes":0.3,"eng":0.15,"save":0.15},"quadrant":"wide","ranks":{"score":{"industry":{"n":71,"of":1162},"company":{"n":6,"of":34}},"views":{"industry":{"n":23,"of":1162},"company":{"n":2,"of":34}},"likes":{"industry":{"n":9,"of":1162},"company":{"n":1,"of":34}},"eng":{"industry":{"n":924,"of":1162},"company":{"n":27,"of":34}},"save":{"industry":{"n":717,"of":1162},"company":{"n":23,"of":34}}},"percentiles":{"industry":{"views":0.9806368330464716,"likes":0.9926850258175559,"eng":0.20524956970740102,"save":0.38339070567986233},"company":{"views":0.9558823529411765,"likes":0.9852941176470589,"eng":0.22058823529411764,"save":0.3382352941176471}},"peers":{"industry":1162,"company":34,"companyName":"Z.ai"},"medians":{"industry":{"views":39945,"likes":289,"eng":0.007331788144862199,"save":0.0007642791742968587},"company":{"views":201224.5,"likes":985.5,"eng":0.005574811911403666,"save":0.0007534579749246171},"form":{"n":108,"views":87497,"eng":0.006289672292921017}},"ratios":{"vsCompanyViews":34.57975544727407,"vsCompanyLikes":24.184677828513443,"viewsPerFollower":42.100800474355175,"viewsPerDay":144522.61010877576},"followers":165277,"reads":[{"at":"2026-10-11T01:27:39.562Z","views":6954648,"likes":23841,"reposts":2614,"quotes":2012,"replies":1002,"saves":4189,"source":"chart-os"},{"at":"2026-10-12T14:53:32.420Z","views":6956698,"likes":23828,"reposts":2614,"quotes":2012,"replies":1002,"saves":4189,"source":"x-search-cache"},{"at":"2026-10-12T20:32:09.336Z","views":6957155,"likes":23831,"reposts":2614,"quotes":2012,"replies":1003,"saves":4189,"source":"x-search-cache"},{"at":"2026-10-12T20:32:35.800Z","views":6957155,"likes":23831,"reposts":2614,"quotes":2012,"replies":1003,"saves":4189,"source":"x-search-cache"},{"at":"2026-10-13T17:43:55.213Z","views":6958294,"likes":23834,"reposts":2615,"quotes":2012,"replies":1002,"saves":4188,"source":"bird-read"}],"chartData":"/api/charts/8f5228bd5d9a.json","copy":{"en":{"why":[{"factor":"reach-vs-company","text":"Its 7M views are 35x Z.ai's median post, ranking #2 of 34 Z.ai posts by views."},{"factor":"views-per-follower","text":"At 42 views per follower, it sits in the top decile and 99th percentile among posts with known followers."},{"factor":"rank","text":"A score of 78 places it #71 of 1,162 AI chart posts and #6 of 34 Z.ai posts."}],"verdict":"It sits in the Wide but thin quadrant, at #71 of 1,162 AI chart posts by score.","news":{"headline":"Z.ai posts six-benchmark LLM chart, 7M views","dek":"The grouped bar chart drew 7M views and 23.8K likes, with an engagement rate of 0.42% and a save rate of 0.060%.","what":"Z.ai posted a grouped bar chart on August 26, 2026, covering LLM performance across six benchmarks. The same day the account posted 2 posts, including one on Agentic Code Bench performance by effort level at 201K views.","numbers":"It reached 7M views, #23 of 1,162 AI chart posts by views, against a median of 39.9K among AI chart posts. Its score of 78 ranks #71 of 1,162 AI chart posts and #6 of 34 Z.ai posts.","lesson":"A single grouped bar chart with 6 models shown drew 7M views, 35x Z.ai's median post. The 0.42% engagement rate and 0.060% save rate sit below the reach, so the chart travelled wide but thin.","implication":"At #6 of 34 Z.ai posts by score and #2 of 34 by views, this chart post is among the account's strongest by reach. The 35x gap to Z.ai's median post shows the account's chart posts can draw views far above its typical post."}},"zh":{"why":[{"factor":"reach-vs-company","text":"696万浏览量是Z.ai公司中位数的35倍，在34条Z.ai帖子中排第2。"},{"factor":"views-per-follower","text":"每粉丝浏览量42，是已知粉丝数帖子中位数0.20的第99百分位。"},{"factor":"rank","text":"综合得分78，在1162条AI图表帖中排第71，在Z.ai的34条帖子中排第6。"}],"verdict":"帖子属于传得远，互动偏少象限，得分在1162条AI图表帖中排第71。","news":{"headline":"Z.ai发布GLM-5.3-Flash多基准图表","dek":"Z.ai在2026年8月26日发布1张分组柱状图，展示6个模型在多个基准上的得分。","what":"Z.ai于2026年8月26日发布1张分组柱状图，展示6个模型在多个基准上的表现。同日该账号另发1条帖子，主题为Agentic Code Bench performance by effort level，浏览量20.1万。","numbers":"帖子浏览量696万，在1162条AI图表帖中排第23，在34条Z.ai的帖子中排第2。得分78，在1162条AI图表帖中排第71。","lesson":"6个模型的分组柱状图拿到696万浏览量，是Z.ai公司中位数20.1万的35倍。图表显示Terminal Bench 2.1的得分从81到87.4，数值密集的对比表是读者保存的对象。","implication":"该帖浏览量排Z.ai第2，但互动率0.42%低于AI图表帖中位数，收藏率0.060%。对Z.ai而言，高浏览量帖的互动率与公司自身中位数之间的差距，是图表发布中可测量的部分。"}},"lane":"amd/DeepSeek-V4.1-Flash","checkedAt":"2026-10-13T19:55:16.155Z"}}