According to CoinMeta, DeepSeek has released the first batch of multi-modal agent scores for v4-flash-vision-exp. In four multi-modal agent evaluations, it achieved a record of 2 wins and 2 losses against Opus: agents ' last exam was 27.3 to 25.7, zerobench was 35.0 to 34.0, apexbench was 36.5 to 39.4, and chartography was 64.3 to 65.0. Compared to the plain-text version v4-flash-0731, the visual version saw an improvement from 26.2 to 36.5 in apexbench, and last ' exam increased from 25.2 to 27.3. It's important to note that these results come from DeepSeek's official self-tests, not from any third-party independent rankings.