Anthropic disclosed the Model2 model for the first time: its performance exceeds the current strongest AI, and the technical boundaries are once again broadened. The field of artificial intelligence has recently ushered in new technological trends.

Anthropic disclosed the Model 2 model for the first time: its performance exceeds the current strongest AI, and the technical boundaries are once again broadened. The field of artificial intelligence has recently ushered in new technological trends. According to the "August 2026 Risk Report" released by Anthropic on August 15, the company disclosed for the first time a new AI model that has not yet been officially released - Model 2. Judging from the details disclosed in the report, Model 2's overall performance on multiple benchmark tasks is better than the currently disclosed industry's strongest model Claude Mythos 5. However, industry analysts pointed out that compared with the previous generation products, the performance increase brought by Model 2 is relatively mild. Its core significance is to further consolidate and improve the overall comprehensive capabilities of the AI ​​​​model, rather than making a disruptive leap. In addition, according to the results of the "CoBench v2" benchmark test within Anthropic, related titles about Model 1 also appeared in the report. Based on this speculation, outsiders speculate that Model 2 is likely to be a subsequent evolved version of Model 1, and both are likely to be further iterations and upgrades of the Claude Mythos Preview technology route. via AI News (author: AI Base)