3027 Commits

Author SHA1 Message Date
fc97232b6f 修复画布已出图取回和任务状态提示 2026-11-17 17:11:37 +08:00
e3f17cbb5d 按固定清晰度线路契约提交 GPT Image 2.6 比例 2026-11-17 13:59:01 +08:00
020637f13d 固定一档的生图线路独立传递所选比例 2026-11-17 13:58:17 +08:00
121ca3014b 修复 GPT Image 2.6 比例与像素尺寸的映射 2026-11-17 13:28:29 +08:00
83091c0457 按像素预设提交独立比例并保留已声明的清晰度 2026-11-17 13:27:31 +08:00
399aa1d761 Merge pull request '修复生图自动比例与固定尺寸冲突' (#100) from codex/image-auto-ratio-20261117 into main 2026-11-17 10:57:50 +08:00
7094231a55 生图自动比例保留清晰度档位 2026-11-17 10:56:06 +08:00
902ad6fc9c Merge pull request '记录语音画布正式上线与网页验收结果' (#99) from codex/daily-model-followup-20261116 into main 2026-11-16 18:55:17 +08:00
bf4dfe0fcb Record production audio canvas acceptance and release status 2026-11-16 18:54:45 +08:00
582667dd05 Merge pull request '保留用户主动清空的画布文字' (#98) from codex/daily-model-followup-20261116 into main 2026-11-16 18:39:43 +08:00
4cfadcea08 Keep intentionally cleared canvas text empty on reload 2026-11-16 18:39:11 +08:00
a7fdff59fc Merge pull request '保留画布恢复后的语音识别文字' (#97) from codex/daily-model-followup-20261116 into main 2026-11-16 18:16:15 +08:00
58b4695d17 Preserve speech recognition text when restoring canvas nodes 2026-11-16 18:15:58 +08:00
2123e6208c Merge pull request '修复语音识别空白结果与音频任务状态显示' (#96) from codex/daily-model-followup-20261116 into main 2026-11-16 17:55:02 +08:00
60c8109d09 修复语音识别结果回填与音频任务状态显示 2026-11-16 17:54:15 +08:00
9687561b9c Merge pull request '记录生图额度故障及充值后的生成恢复' (#95) from codex/chat-image-fallback-20261116 into main 2026-11-16 16:06:51 +08:00
82218217ec 记录模型参数修复上线结果 2026-11-16 16:06:47 +08:00
2d65834dbb 补充模型参数故障的线上核对与恢复记录 2026-11-16 15:57:17 +08:00
0a808a2b83 Merge pull request '修复生图额度错误分类和备用模型参数不匹配' (#94) from codex/chat-image-fallback-20261116 into main 2026-11-16 15:36:34 +08:00
90492972b3 修复生图额度错误分类与备用模型规格筛选 2026-11-16 15:36:20 +08:00
4d739927f7 Merge pull request '修复生效模型的参数映射与能力配置不一致' (#93) from codex/model-parameter-audit-20261116 into main 2026-11-16 14:43:59 +08:00
903686db36 fix: align enabled model parameters with provider contracts 2026-11-16 14:43:24 +08:00
40fdb41b32 Merge pull request '修复 4K 图片请求的尺寸与清晰度不一致' (#92) from codex/98k-image-resolution-20261116 into main 2026-11-16 13:30:02 +08:00
f980cfb036 fix: keep 4K image dimensions consistent with resolution 2026-11-16 13:29:04 +08:00
2d0238aca6 Merge pull request '参考图编辑按公开合同发送 JSON 和表单请求' (#91) from codex/98k-reference-edits-20261116 into main 2026-11-16 12:20:28 +08:00
5701e95bb4 fix: encode published image editing contracts correctly 2026-11-16 12:19:53 +08:00
36ad7900b6 Merge pull request '首帧视频比例遵循模型公开约束' (#90) from codex/98k-standard-followup-20261116 into main 2026-11-16 11:54:58 +08:00
c655bbd81e fix: honor published first-frame video aspect ratios 2026-11-16 11:51:57 +08:00
dff2e22ffd Merge pull request '生成视频由七牛直接抓取并校验可用性' (#89) from codex/98k-model-expansion-20261116 into main 2026-11-16 11:13:22 +08:00
995820138a perf: archive generated videos directly through Qiniu 2026-11-16 11:12:52 +08:00
4ac5062080 Merge pull request '适配 98K 图片、视频、对话及音频模型' (#88) from codex/98k-model-expansion-20261116 into main 2026-11-16 11:00:16 +08:00
ba24efc8e7 feat: support 98K image, video, chat and audio model contracts 2026-11-16 10:57:53 +08:00
1a682343c7 Merge pull request '转换后的下载文件优先使用七牛表单上传' (#87) from codex/direct-download-archive-20261113 into main 2026-11-13 18:45:51 +08:00
f8c97ba7e4 fix: use Qiniu form uploads for prepared download files 2026-11-13 18:45:32 +08:00
17075e38bd Merge pull request '原图下载等待七牛准备并拆分图片存储队列' (#86) from codex/direct-download-archive-20261113 into main 2026-11-13 18:40:04 +08:00
ed25603be2 fix: prepare CDN originals without download relay fallback 2026-11-13 18:39:43 +08:00
7464108d2f Merge pull request '对齐98K异步等待并保留待核对任务编号' (#85) from codex/image-review-requirements-20261113 into main 2026-11-13 15:02:50 +08:00
86fd3353f4 对齐98K异步等待并保留待核对任务编号 2026-11-13 15:02:45 +08:00
af54ec2611 Merge pull request '按用户输出要求验收图片并返回识图内容' (#84) from codex/image-review-requirements-20261113 into main 2026-11-13 14:49:27 +08:00
9139b33208 按用户输出要求验收图片并返回识图内容 2026-11-13 14:49:21 +08:00
2b5455d046 Merge pull request '明确提示不支持的生图清晰度尚未提交' (#83) from codex/wutong-image-submission-20261113 into main 2026-11-13 10:14:36 +08:00
89a9416815 fix: report unsupported image quality before submission 2026-11-13 10:14:31 +08:00
168de1db3c Merge pull request '修复自动生图模型选择与提交结果确认' (#82) from codex/wutong-image-submission-20261113 into main 2026-11-13 09:58:14 +08:00
90fe4f2933 fix: align image routing, submission recovery and output quality 2026-11-13 09:58:09 +08:00
dc7a1ae0a4 Merge pull request '七牛侧准备 WebP 兼容下载并保留色彩配置' (#81) from codex/qiniu-png-preparation-20261112 into main 2026-11-12 16:44:21 +08:00
a6f79998bd Prepare uploaded WebP downloads in Qiniu with embedded colour profiles 2026-11-12 16:43:52 +08:00
d0653c2d51 Merge pull request '补齐图片直传七牛、CDN 展示和兼容原图下载' (#80) from codex/qiniu-direct-delivery-20261112 into main 2026-11-12 16:21:27 +08:00
f058012bb8 Complete direct Qiniu media upload, display and download paths 2026-11-12 16:20:56 +08:00
8aa9ebd061 Merge pull request '扩图结果直接归档到七牛存储' (#79) from codex/native-expansion-archive-20261112 into main 2026-11-12 15:09:46 +08:00
d32258e13b 扩图结果直接归档到七牛存储 2026-11-12 15:09:33 +08:00