v0.9.1
更新内容 / Updates
-
生图功能继续完善:新增生图批次协调、并发控制、短期幂等复用和更可靠的结果交接。
-
Continued improving image generation with batch coordination, concurrency control, short-lived idempotent reuse, and more reliable result handoff.
-
增强 Codex Responses 生图工具桥接、图像结果解析、预览压缩和安全下载能力。
-
Enhanced the Codex Responses image-tool bridge, image-result parsing, preview compression, and secure downloads.
-
增加生图后端配置、模型开关、请求日志、统计和管理界面支持。
-
Added image-backend configuration, per-model switches, request logs, statistics, and admin UI support.
-
改进生产启动、模型发现、登录限流、请求正文脱敏及容器/绿色版发布配置。
-
Improved production startup, model discovery, login throttling, request-payload redaction, and container/green-package release configuration.
-
这是围绕生图能力的兼容性和稳定性小版本;长程生图任务仍可能存在限制,后续会继续完善。
-
This is a compatibility and stability patch release centered on image generation; long-running image-generation tasks may still have limitations and will continue to be improved.
验证 / Verification
- JavaScript syntax:
node --check app/web/static/app.jspassed. - Full test suite:
602 passed. - GitHub Actions test, package, release, and GHCR jobs completed successfully.
- Windows, Linux, and macOS ARM64 packages uploaded successfully.
Docker image:
docker pull ghcr.io/pricening/llm-aio-gateway:latest