{"id":57643,"date":"2026-09-17T09:27:35","date_gmt":"2026-09-17T01:27:35","guid":{"rendered":"https:\/\/www.1ai.net\/?p=57643"},"modified":"2026-09-17T09:27:35","modified_gmt":"2026-09-17T01:27:35","slug":"mozilla-%e6%8a%a5%e5%91%8a%ef%bc%9a%e4%b8%ad%e7%be%8e-ai-%e6%a8%a1%e5%9e%8b%e8%83%bd%e5%8a%9b%e5%b7%ae%e8%b7%9d%e7%bc%a9%e7%9f%ad%e8%87%b3-4-4-%e4%b8%aa%e6%9c%88","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/57643.html","title":{"rendered":"Mozilla Report: CUSIC AI model capability gap reduced to 4.4 months"},"content":{"rendered":"

The news of September 17thMozilla<\/a> (15 SEPTEMBER) ISSUED THE SECOND EDITION OF THE OPEN SOURCE AI STATUS REPORT, WHICH REPORTED THATChina<\/a>Best open weight model of the company andUSA<\/a>The technology company's front-line closed-source modelThe capacity gap has narrowed to 4.4 months\u3002<\/strong><\/p>\n

\"84eb53ecj00tlhina00ild0014,000lum\"<\/p>\n

COMPARED WITH THE FIRST EDITION OF THE OPEN SOURCE AI STATUS REPORT, RELEASED IN JULY OF THIS YEAR, THE SECOND EDITION OF THE REPORT FURTHER LOWERS MACROJUDGMENT TO SPECIFIC MODELS, SPECIFIC MISSION DURATION AND SINGLE TASK COST\u3002<\/p>\n\n\n\n\n\n\n\n
dimension (math.)<\/th>\nJuly first edition<\/th>\nSeptember new edition<\/th>\nMain changes<\/th>\n<\/tr>\n
Capacity gap<\/td>\nThe performance gap between top open models and proprietary models like ChatGPT and Claude is 3%<\/td>\nThe gap between the US front closed source model and the best open weight model in China is 4.4 months<\/td>\nChange from a single, integrated percentage to a closer \u201ctask time\u201d indicator for actual work\u3002<\/td>\n<\/tr>\n
Cost presentation<\/td>\nGPT-4 GRADE REASONING COSTS DROPPED TO ONE IN 50 IN 3 YEARS<\/td>\nKimi K3 cost 30% of Fable 5; GM 5.2 single task cost is about one fifth lower<\/td>\nThe new edition gives a comparison of the model with the direct cost of the mission\u3002<\/td>\n<\/tr>\n
Use data<\/td>\n79% DEVELOPER USES OPEN MODEL 51% DEPLOYED PRODUCTION; CLOSED SOURCE MODEL PRODUCTION DEPLOYMENT RATE 63%<\/td>\nAugust 2026 OpenRouter 8 of 10 models by token provide open weights<\/td>\nThe focus has shifted from research by developers to the latest platform use ranking\u3002<\/td>\n<\/tr>\n
Income data<\/td>\nOPEN MODEL COVERS ABOUT ONE THIRD OF THE ACTUAL AI USAGE, BUT ONLY 4% INCOME<\/td>\nREAFFIRMING THAT OPEN MODEL INCOME IS 4%, CLOSED SOURCE 96%, AND INDICATING THAT DATA COVER MAY TO SEPTEMBER 2025<\/td>\nThe new version complements the time frame of the income conclusion and the share of closed-source income\u3002<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

The second edition focused on the Kimi K3 model on the dark side of the month, which scored only three points lower on the Artificial Analysis smart index than the Anthropic Fable 5, at a cost of about 301 TP3T\u3002<\/p>\n

\"d0cdd158jlhina00erd0014,000m2m\"<\/p>\n

IN TERMS OF SCOPE OF CAPACITY, METR MEASURES THE \u201cTIME HORIZON\u201d OF A MODEL OF A RELIABLE TIME FRAME FOR A MISSION, AND THE RELIABLE SUCCESS RATE IS 50%. CURRENTLY, THE OPTIMAL CLOSED-SOURCE MODEL IS 1.7 TIMES LONGER THAN THE BEST OPEN-SOURCE MODEL. THE OPEN-SOURCE MODEL HANDLES 7 HOURS OF TASK AND THE CLOSED-SOURCE MODEL HANDLES ABOUT 12 HOURS OF TASK\u3002<\/p>\n

\"#cab84e7dj00tlhina00hkd0014,000lym\"<\/p>\n

Mozilla Chief Technical Officer Raffi Krikorian characterized the difference as depending on the mission scenario: the closed-source model premium was mainly on professional tasks that required between 8 and 12 hours, while open-source models were able to handle daily work at very low prices\u3002<\/p>\n

Krikorian states that the open model can handle 12 hours of mission after 4 months and the closed source model can handle approximately 20 hours of mission. At present, 8 to 12 hours of work are usually still areas where closed-source models can be completed and open models are difficult to handle\u3002<\/p>\n

This shift has begun to reshape the behaviour of enterprises. DoorDash is now using Kimi to handle day-to-day tasks while retaining Fable for more complex tasks. In August 2026, eight of the top 10 models on the routed platform OpenRouter ranked by token usage were open source models\u3002<\/p>","protected":false},"excerpt":{"rendered":"

On September 17, Mozilla (15 September) released the second edition of the Open Source AI status report, which reported that the capability gap had narrowed to 4.4 months for the best open weight model of Chinese companies and the front-end closed-source model of American technology companies. Compared with the first edition of the Open Source AI status report, released in July of this year, the second edition of the report further lowers macrojudgment to specific models, specific mission duration and single task cost. The difference between the performance of top-level open models and proprietary models like ChatGPT, Claude, is 3%, the difference between the U.S. front-line closed-source model and China ' s best open-weight model is 4.4 months from a single integrated energy percentage<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[167,406,4159,236],"collection":[],"class_list":["post-57643","post","type-post","status-publish","format-standard","hentry","category-news","tag-ai","tag-mozilla","tag-4159","tag-236"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/57643","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=57643"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/57643\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=57643"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=57643"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=57643"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=57643"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}