Ideogram 4.5 连续 5 次编辑后人脸一致性登顶 Rapidata 基准
编辑图片时人脸最容易越改越走样，这次是拿 199 张照片各改 5 轮、靠 32 万多条人工投票排出来的，Ideogram 4.5 在 20 个模型里拿了第一。做头像、电商图、换装这类反复编辑的活儿，可以照着榜单选模型。
Ideogram 4.5 在 Rapidata 的人脸身份保持基准中排名第一，参评模型共 20 个。该基准对 199 张照片各执行 5 次连续编辑，再评估结果中的人脸是否仍与原图为同一人。基准基于 322K+ 条人类判断，全部输出、判断与方法论已在 Benchmark AI 公开。
测试针对的核心问题是身份漂移：图像编辑模型在多轮修改后常让人物面孔逐渐走样。Ideogram 4.5 在 5 次连续编辑后仍保持人脸一致，取得 20 个模型中的首位。
对人像修图、电商素材、虚拟形象等需要反复编辑同一人物的场景，多轮编辑后的人脸保真是实际选型的关键指标。Rapidata 公开的方法论和 322K+ 人类判断数据可供开发者自行复核各模型表现。
Edit the photo. Keep the person. Ideogram 4.5 ranks #1 among 20 models for preserving facial identity after five consecutive edits in @RapidataAI ’s benchmark. Rapidata @RapidataAI After 5 edits, does the person still look the same? Ideogram 4.5 ranks #1 among 20 models for preserving facial identity! 🥇 We benchmarked character identity drift across 199 photos, applying 5 consecutive edits and evaluating only the resulting faces. This benchmark holds 322K+ human judgments. All outputs, judgments and methodology are publicly available on Benchmark AI.
This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include Ideogram 4.5, Rapidata, 图像编辑, 多模态.
Source attribution: curated from Ideogram (@ideogram_ai).
