开源视频生成模型 Wan2.7-260612 表现出色，在盲人偏好测试中以 81-82 ELO 分的微弱差距落后于 Google 的 Gemini Omni Flash。这表明 Wan2.7-260612 是用户可以下载并在本地运行的领先模型，在竞争激烈的视频生成领域展现出强大的能力。 AI
影响 这一接近的 ELO 分数表明，开源视频生成模型正在迅速接近领先的专有系统的能力。
排序理由 该集群报告了一个开源模型与商业模型相比的基准测试结果，表明了一个研究里程碑。
在 Mastodon — fosstodon.org 阅读 →
AI 生成摘要 · Google Gemini · 来自 5 个来源。 我们如何撰写摘要 →
🎨 Wan2.7-260612 is the best downloadable video model, just 81 ELO behind Google’s Gemini Omni Flash. Open-source is nearly matching commercial — see the full leaderboard here. https:// olud.ai/media-leaderboard.html # OpenSource # GenerativeAI # StableDiffusion # AI
🎨 Where open-source stands in video generation: #6 Wan2.7-260612 — the best model you can actually download 81 ELO points behind Gemini Omni Flash (Google) Blind human preference, not marketing → https:// olud.ai/media-leaderboard.html # OpenSource # GenerativeAI # StableDiffusio…
The best open-source video model you can run locally, Wan2.7-260612, trails Gemini Omni Flash by just 81 ELO in blind human preference. https:// olud.ai/media-leaderboard.html # OpenSource # GenerativeAI # StableDiffusion # AI
🎨 Where open-source stands in video generation: #6 Wan2.7-260612 — the best model you can actually download 82 ELO points behind Gemini Omni Flash (Google) Blind human preference, not marketing → https:// olud.ai/media-leaderboard.html # OpenSource # GenerativeAI # StableDiffusio…
🎨 Wan2.7-260612 leads open-source video generation, just 82 ELO points behind Gemini Omni Flash in blind human preference tests. https:// olud.ai/media-leaderboard.html # OpenSource # GenerativeAI # StableDiffusion # AI