Name: 图像编辑模型排行榜（single image edit photorealistic open source模型排行榜）最好的 AI 图像编辑模型对比 |AITNT
Creator: LM Arena
License: https://creativecommons.org/licenses/by/4.0/
Keywords: single_image_edit_photorealistic_open_source,AI 图像编辑模型排行榜,图像编辑 AI 排行榜,最佳模型对比,AI 模型排行

图像编辑模型 Arena 榜单LM Arena Global AI Leaderboardsingleimage edit photorealistic open sourcesingleimage edit photorealistic open source

比较并探索基于 single_image_edit_photorealistic_open_source 场景进行 Elo 积分排名的主流 AI 图像编辑模型，数据来源于真实用户匿名盲测。Explore Elo rankings of AI image editing models for single and multi-image edit tasks — rated by real anonymous user votes.

2026-10-06

▾

上榜模型Models

参赛厂商Providers

432.0K

总投票数Total Votes

📖 指标说明 — 读懂这张榜单📖 Metrics Guide — Understanding the Leaderboard

🎯

榜单主指标（Score）Primary Metric (Score)

每个榜单会采用官网对应的核心排序指标。通用 Arena 榜单通常是模型竞技积分；智能体榜单则展示净提升及多项任务指标。

Each leaderboard uses the official primary ranking metric. Most Arena boards use competitive scores; the Agent board uses net improvement and task-specific metrics.

🔢

精确值（Float Value）Float Value

主指标的完整浮点数值。对于通用 Arena 榜单可用于区分整数分相近的模型；智能体榜单以百分比和置信区间展示。

The full floating-point value of the primary metric. It helps distinguish close scores on standard boards; Agent metrics are shown as percentages with confidence intervals.

📊

方差（Variance）Variance

衡量结果的统计稳定程度。绿色·稳定（<5）更可信；橙色·波动（5~12）；红色·不稳（>12）说明排名还可能明显变化。

Measures statistical stability. Green·Stable (<5); Orange·Unstable (5–12); Red·Volatile (>12) means the ranking may still move significantly.

↔️

置信区间（95% CI = ±N）95% Confidence Interval (CI = ±N)

95% 置信区间反映当前估计的不确定范围，显示为 ±N。在相同计算口径下，N 越小说明估计越集中、排名越稳定。

The 95% confidence interval, shown as ±N, summarizes uncertainty in the current estimate. Under the same methodology, a smaller N indicates a more precise estimate and a more stable ranking.

🗳️

投票数（Vote Count）Vote Count

该模型或厂商参与评测的总次数。数量越高，结果的统计置信度越高、越不容易受单次样本影响而波动。

Total evaluation volume for the model or lab. Higher counts mean higher statistical confidence and less sensitivity to individual samples.

📜

开源协议（License）License

Apache/Llama：完全开源可商用；MIT：开源限制极少；Proprietary：闭源商业模型。协议决定你能否把它集成到自己的产品里。

Apache/Llama: Fully open-source, commercial use allowed; MIT: Open-source, minimal restrictions; Proprietary: Closed-source commercial model. License determines if you can integrate it into your product.

排名Rank	名次区间Rank Range	模型Model		投票数Votes	精确分 & 区间Float Score & CI	方差Variance	价格 ($/1M)Price ($/1M)	上下文Context
🥇	15 - 17	Qwen-Image-2.1 阿里巴巴 · qwen-research	1364±6.0Preliminary	13.7K 票votes	1364.000 [1358.0, 1370.0]	2.00 · 稳定Stable	—	—
🥈	22 - 26	HunyuanImage 3.0 Instruct 腾讯 · tencent-hunyuan-community	1311±5.0	84.6K 票votes	1311.000 [1306.0, 1316.0]	4.00 · 稳定Stable	—	—
🥉	30 - 32	Qwen-Image-Edit 阿里巴巴 · Apache 2.0	1255±5.0	38.2K 票votes	1255.000 [1250.0, 1260.0]	2.00 · 稳定Stable	—	—
4	34 - 36	Qwen-Image-Edit (2511) 阿里巴巴 · Apache 2.0	1233±4.0	119.1K 票votes	1233.000 [1229.0, 1237.0]	2.00 · 稳定Stable	—	—
5	34 - 36	FLUX.2 [klein] 9B Black Forest Labs · flux-non-commercial-license	1229±5.0	62.1K 票votes	1229.000 [1224.0, 1234.0]	2.00 · 稳定Stable	—	—
6	37 - 41	FLUX.2 [dev] Black Forest Labs · flux-non-commercial-license	1213±5.0	52.0K 票votes	1213.000 [1208.0, 1218.0]	4.00 · 稳定Stable	—	—
7	42 - 42	FLUX.2 [klein] 4B Black Forest Labs · Apache 2.0	1189±5.0	62.3K 票votes	1189.000 [1184.0, 1194.0]	0.00 · 稳定Stable	—	—

🔍

没有找到匹配的模型，换个关键词试试？No matching models found. Try a different keyword?